+ [MAJOR] [ROOT] [CREATE] 1. fork repo from COVER github

.gitignore

@@ -0,0 +1,164 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+pip-wheel-metadata/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+.python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+.idea/
+
+# VSCode
+.vscode/
+
+# JupyterLab
+.jupyterlab/
+
+# Data directories (you might want to keep them in version control, depending on your project)
+#/data
+#/models
+
+# Log files
+*.log
+
+# Others
+.DS_Store
+.pretrained_weights/*
+.datasets/
+datasets/KoNViD
+datasets/LIVE_VQC
+datasets/YouTubeUGC
+*.pth
+*.swp
+*.bak
+*.tmp
+*.temp

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Zhengzhong Tu
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README copy.md

@@ -0,0 +1,163 @@
+# COVER
+
+Official Code for [CVPR Workshop2024] Paper *"COVER: A Comprehensive Video Quality Evaluator"*. 
+Official Code, Demo, Weights for the [Comprehensive Video Quality Evaluator (COVER)].
+
+# Todo:: update date, hugging face model below
+- xx xxx, 2024: We upload weights of [COVER](https://github.com/vztu/COVER/release/Model/COVER.pth) and [COVER++](TobeContinue) to Hugging Face models.
+- xx xxx, 2024: We upload Code of [COVER](https://github.com/vztu/COVER)
+- 12 Apr, 2024: COVER has been accepted by CVPR Workshop2024.
+
+
+# Todo:: update [visitors](link) below
+![visitors](https://visitor-badge.laobi.icu/badge?page_id=teowu/TobeContinue) [![](https://img.shields.io/github/stars/vztu/COVER)](https://github.com/vztu/COVER)
+[![State-of-the-Art](https://cdn.rawgit.com/sindresorhus/awesome/d7305f38d29fed78fa85652e3a63e154dd8e8829/media/badge.svg)](https://github.com/QualityAssessment/COVER)
+<a href="https://colab.research.google.com/github/taskswithcode/COVER/blob/master/TWCCOVER.ipynb"><img src="https://colab.research.google.com/assets/colab-badge.svg" alt="google colab logo"></a> 
+
+
+# Todo:: update predicted score for YT-UGC challenge dataset specified by AIS
+**COVER** Pseudo-labelled Quality scores of [YT-UGC](https://www.deepmind.com/open-source/kinetics): [CSV](https://github.com/QualityAssessment/COVER/raw/master/cover_predictions/kinetics_400_1.csv)
+
+
+[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/disentangling-aesthetic-and-technical-effects/video-quality-assessment-on-youtube-ugc)](https://paperswithcode.com/sota/video-quality-assessment-on-youtube-ugc?p=disentangling-aesthetic-and-technical-effects)
+
+
+## Introduction
+#  Todo:: Add Introduction here
+
+### the proposed COVER
+
+*This inspires us to*
+
+![Fig](figs/approach.png)
+
+## Install
+
+The repository can be installed via the following commands:
+```shell
+git clone https://github.com/vztu/COVER 
+cd COVER 
+pip install -e . 
+mkdir pretrained_weights 
+cd pretrained_weights 
+wget https://github.com/vztu/COVER/release/Model/COVER.pth
+cd ..
+```
+
+
+## Evaluation: Judge the Quality of Any Video
+
+### Try on Demos
+You can run a single command to judge the quality of the demo videos in comparison with videos in VQA datasets. 
+
+```shell
+    python evaluate_one_video.py -v ./demo/video_1.mp4
+```
+
+or 
+
+```shell
+    python evaluate_one_video.py -v ./demo/video_2.mp4
+```
+
+Or choose any video you like to predict its quality:
+
+
+```shell
+    python evaluate_one_video.py -v $YOUR_SPECIFIED_VIDEO_PATH$
+```
+
+### Outputs
+
+#### ITU-Standarized Overall Video Quality Score
+
+The script can directly score the video's overall quality (considering all perspectives).
+
+```shell
+    python evaluate_one_video.py -v $YOUR_SPECIFIED_VIDEO_PATH$
+```
+
+The final output score is averaged among all perspectives.
+
+
+## Evaluate on a Exsiting Video Dataset
+
+
+```shell
+    python evaluate_one_dataset.py -in $YOUR_SPECIFIED_DIR$ -out $OUTPUT_CSV_PATH$
+```
+
+## Evaluate on a Set of Unlabelled Videos
+
+
+```shell
+    python evaluate_a_set_of_videos.py -in $YOUR_SPECIFIED_DIR$ -out $OUTPUT_CSV_PATH$
+```
+
+The results are stored as `.csv` files in cover_predictions in your `OUTPUT_CSV_PATH`.
+
+Please feel free to use COVER to pseudo-label your non-quality video datasets.
+
+
+## Data Preparation
+
+We have already converted the labels for most popular datasets you will need for Blind Video Quality Assessment,
+and the download links for the **videos** are as follows:
+
+:book: LSVQ: [Github](https://github.com/baidut/PatchVQ)
+
+:book: KoNViD-1k: [Official Site](http://database.mmsp-kn.de/konvid-1k-database.html)
+
+:book: LIVE-VQC: [Official Site](http://live.ece.utexas.edu/research/LIVEVQC)
+
+:book: YouTube-UGC: [Official Site](https://media.withyoutube.com)
+
+*(Please contact the original authors if the download links were unavailable.)*
+
+After downloading, kindly put them under the `../datasets` or anywhere but remember to change the `data_prefix` respectively in the [config file](cover.yml).
+
+# Training: Adapt COVER to your video quality dataset!
+
+Now you can employ ***head-only/end-to-end transfer*** of COVER to get dataset-specific VQA prediction heads. 
+
+We still recommend **head-only** transfer. As we have evaluated in the paper, this method has very similar performance with *end-to-end transfer* (usually 1%~2% difference), but will require **much less** GPU memory, as follows:
+
+```shell
+    python transfer_learning.py -t $YOUR_SPECIFIED_DATASET_NAME$
+```
+
+For existing public datasets, type the following commands for respective ones:
+
+- `python transfer_learning.py -t val-kv1k` for KoNViD-1k.
+- `python transfer_learning.py -t val-ytugc` for YouTube-UGC.
+- `python transfer_learning.py -t val-cvd2014` for CVD2014.
+- `python transfer_learning.py -t val-livevqc` for LIVE-VQC.
+
+
+As the backbone will not be updated here, the checkpoint saving process will only save the regression heads with only `398KB` file size (compared with `200+MB` size of the full model). To use it, simply replace the head weights with the official weights [COVER.pth](https://github.com/vztu/COVER/release/Model/COVER.pth).
+
+We also support ***end-to-end*** fine-tune right now (by modifying the `num_epochs: 0` to `num_epochs: 15` in `./cover.yml`). It will require more memory cost and more storage cost for the weights (with full parameters) saved, but will result in optimal accuracy.
+
+Fine-tuning curves by authors can be found here: [Official Curves](https://wandb.ai/timothyhwu/COVER) for reference.
+
+
+## Visualization
+
+### WandB Training and Evaluation Curves
+
+You can be monitoring your results on WandB!
+
+## Acknowledgement
+
+Thanks for every participant of the subjective studies!
+
+## Citation
+
+Should you find our work interesting and would like to cite it, please feel free to add these in your references! 
+
+
+# Todo, add bibtex of cover below
+```bibtex
+%cover
+
+```

_config.yaml

@@ -0,0 +1,25 @@
+theme: minima
+
+
+encoding: "utf-8"
+markdown_ext: "markdown,mkdown,mkdn,mkd,md"
+
+
+# Conversion
+markdown: kramdown
+highlighter: rouge
+lsi: false
+excerpt_separator: "\n\n"
+incremental: false
+
+
+# Markdown Processing
+kramdown:
+  input: GFM
+  hard_wrap: false
+  auto_ids: true
+  footnote_nr: 1
+  entity_output: as_char
+  toc_levels: 1..6
+  smart_quotes: lsquo,rsquo,ldquo,rdquo
+  enable_coderay: false

cover.yml

@@ -0,0 +1,236 @@
+name: COVER
+num_epochs: 0
+l_num_epochs: 10
+warmup_epochs: 2.5
+ema: true
+save_model: true
+batch_size: 8
+num_workers: 6
+split_seed: 42
+
+wandb:
+    project_name: COVER
+
+data:   
+    val-livevqc:
+        type: ViewDecompositionDataset
+        args:
+            weight: 0.598
+            phase: test
+            anno_file: ./examplar_data_labels/LIVE_VQC/labels.txt
+            data_prefix: ./datasets/LIVE_VQC/ # revert before submit
+            sample_types:
+                semantic:
+                    size_h: 512
+                    size_w: 512
+                    clip_len: 20
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+                technical:
+                    fragments_h: 7
+                    fragments_w: 7
+                    fsize_h: 32
+                    fsize_w: 32
+                    aligned: 40
+                    clip_len: 40
+                    t_frag: 20
+                    frame_interval: 2
+                    num_clips: 1
+                aesthetic:
+                    size_h: 224
+                    size_w: 224
+                    clip_len: 40
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+                    
+    val-kv1k:
+        type: ViewDecompositionDataset
+        args:
+            weight: 0.540
+            phase: test
+            anno_file: ./examplar_data_labels/KoNViD/labels.txt
+            data_prefix: ./datasets/KoNViD/ # revert before submit
+            sample_types:
+                semantic:
+                    size_h: 512
+                    size_w: 512
+                    clip_len: 20
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+                technical:
+                    fragments_h: 7
+                    fragments_w: 7
+                    fsize_h: 32
+                    fsize_w: 32
+                    aligned: 40
+                    clip_len: 40
+                    t_frag: 20
+                    frame_interval: 2
+                    num_clips: 1
+                aesthetic:
+                    size_h: 224
+                    size_w: 224
+                    clip_len: 40
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+
+    val-ltest:
+        type: ViewDecompositionDataset
+        args:
+            weight: 0.603
+            phase: test
+            anno_file: ./examplar_data_labels/LSVQ/labels_test.txt
+            data_prefix: ./datasets/LSVQ/ # revert before submit
+            sample_types:
+                semantic:
+                    size_h: 512
+                    size_w: 512
+                    clip_len: 20
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+                technical:
+                    fragments_h: 7
+                    fragments_w: 7
+                    fsize_h: 32
+                    fsize_w: 32
+                    aligned: 40
+                    clip_len: 40
+                    t_frag: 20
+                    frame_interval: 2
+                    num_clips: 1
+                aesthetic:
+                    size_h: 224
+                    size_w: 224
+                    clip_len: 40
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+
+    val-l1080p:
+        type: ViewDecompositionDataset
+        args:
+            weight: 0.620
+            phase: test
+            anno_file: ./examplar_data_labels/LSVQ/labels_1080p.txt
+            data_prefix: ./datasets/LSVQ/ # revert before submit
+            sample_types:
+                semantic:
+                    size_h: 512
+                    size_w: 512
+                    clip_len: 20
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+                technical:
+                    fragments_h: 7
+                    fragments_w: 7
+                    fsize_h: 32
+                    fsize_w: 32
+                    aligned: 40
+                    clip_len: 40
+                    t_frag: 20
+                    frame_interval: 2
+                    num_clips: 1
+                aesthetic:
+                    size_h: 224
+                    size_w: 224
+                    clip_len: 40
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+
+    val-cvd2014:
+        type: ViewDecompositionDataset
+        args:
+            weight: 0.576
+            phase: test
+            anno_file: ./examplar_data_labels/CVD2014/labels.txt
+            data_prefix: ./datasets/CVD2014/ # revert before submit
+            sample_types:
+                semantic:
+                    size_h: 512
+                    size_w: 512
+                    clip_len: 20
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+                technical:
+                    fragments_h: 7
+                    fragments_w: 7
+                    fsize_h: 32
+                    fsize_w: 32
+                    aligned: 40
+                    clip_len: 40
+                    t_frag: 20
+                    frame_interval: 2
+                    num_clips: 1
+                aesthetic:
+                    size_h: 224
+                    size_w: 224
+                    clip_len: 40
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+
+    val-ytugc:
+        type: ViewDecompositionDataset
+        args:
+            weight: 0.443
+            phase: test
+            anno_file: ./examplar_data_labels/YouTubeUGC/labels.txt
+            data_prefix: ./dataset/YouTubeUGC/ # revert before submit
+            sample_types:
+                semantic:
+                    size_h: 512
+                    size_w: 512
+                    clip_len: 20
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+                technical:
+                    fragments_h: 7
+                    fragments_w: 7
+                    fsize_h: 32
+                    fsize_w: 32
+                    aligned: 40
+                    clip_len: 40
+                    t_frag: 20
+                    frame_interval: 2
+                    num_clips: 1
+                aesthetic:
+                    size_h: 224
+                    size_w: 224
+                    clip_len: 40
+                    frame_interval: 2
+                    t_frag: 20
+                    num_clips: 1
+
+model:
+    type: COVER
+    args:
+        backbone:
+            technical:
+                type: swin_tiny_grpb
+                checkpoint: true
+                pretrained:
+            aesthetic:
+                type: conv_tiny
+            semantic:
+                type: clip_iqa+
+        backbone_preserve_keys: technical,aesthetic,semantic
+        divide_head: true
+        vqa_head:
+            in_channels: 768
+            hidden_channels: 64
+            
+optimizer:
+    lr: !!float 1e-3
+    backbone_lr_mult: !!float 1e-1
+    wd: 0.05
+
+test_load_path: ./pretrained_weights/COVER.pth # revert before submit

cover/__init__.py

@@ -0,0 +1,2 @@
+from .datasets import *
+from .models import *

cover/datasets/__init__.py

@@ -0,0 +1,3 @@
+## API for COVER and its variants
+from .basic_datasets import *
+from .cover_datasets import *

cover/datasets/basic_datasets.py

@@ -0,0 +1,812 @@
+import os.path as osp
+import random
+
+import cv2
+import decord
+import numpy as np
+import skvideo.io
+import torch
+import torchvision
+from decord import VideoReader, cpu, gpu
+from tqdm import tqdm
+
+random.seed(42)
+
+decord.bridge.set_bridge("torch")
+
+
+def get_spatial_fragments(
+    video,
+    fragments_h=7,
+    fragments_w=7,
+    fsize_h=32,
+    fsize_w=32,
+    aligned=32,
+    nfrags=1,
+    random=False,
+    fallback_type="upsample",
+):
+    size_h = fragments_h * fsize_h
+    size_w = fragments_w * fsize_w
+
+    ## situation for images
+    if video.shape[1] == 1:
+        aligned = 1
+
+    dur_t, res_h, res_w = video.shape[-3:]
+    ratio = min(res_h / size_h, res_w / size_w)
+    if fallback_type == "upsample" and ratio < 1:
+
+        ovideo = video
+        video = torch.nn.functional.interpolate(
+            video / 255.0, scale_factor=1 / ratio, mode="bilinear"
+        )
+        video = (video * 255.0).type_as(ovideo)
+
+    assert dur_t % aligned == 0, "Please provide match vclip and align index"
+    size = size_h, size_w
+
+    ## make sure that sampling will not run out of the picture
+    hgrids = torch.LongTensor(
+        [min(res_h // fragments_h * i, res_h - fsize_h) for i in range(fragments_h)]
+    )
+    wgrids = torch.LongTensor(
+        [min(res_w // fragments_w * i, res_w - fsize_w) for i in range(fragments_w)]
+    )
+    hlength, wlength = res_h // fragments_h, res_w // fragments_w
+
+    if random:
+        print("This part is deprecated. Please remind that.")
+        if res_h > fsize_h:
+            rnd_h = torch.randint(
+                res_h - fsize_h, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_h = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+        if res_w > fsize_w:
+            rnd_w = torch.randint(
+                res_w - fsize_w, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_w = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+    else:
+        if hlength > fsize_h:
+            rnd_h = torch.randint(
+                hlength - fsize_h, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_h = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+        if wlength > fsize_w:
+            rnd_w = torch.randint(
+                wlength - fsize_w, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_w = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+
+    target_video = torch.zeros(video.shape[:-2] + size).to(video.device)
+    # target_videos = []
+
+    for i, hs in enumerate(hgrids):
+        for j, ws in enumerate(wgrids):
+            for t in range(dur_t // aligned):
+                t_s, t_e = t * aligned, (t + 1) * aligned
+                h_s, h_e = i * fsize_h, (i + 1) * fsize_h
+                w_s, w_e = j * fsize_w, (j + 1) * fsize_w
+                if random:
+                    h_so, h_eo = rnd_h[i][j][t], rnd_h[i][j][t] + fsize_h
+                    w_so, w_eo = rnd_w[i][j][t], rnd_w[i][j][t] + fsize_w
+                else:
+                    h_so, h_eo = hs + rnd_h[i][j][t], hs + rnd_h[i][j][t] + fsize_h
+                    w_so, w_eo = ws + rnd_w[i][j][t], ws + rnd_w[i][j][t] + fsize_w
+                target_video[:, t_s:t_e, h_s:h_e, w_s:w_e] = video[
+                    :, t_s:t_e, h_so:h_eo, w_so:w_eo
+                ]
+    # target_videos.append(video[:,t_s:t_e,h_so:h_eo,w_so:w_eo])
+    # target_video = torch.stack(target_videos, 0).reshape((dur_t // aligned, fragments, fragments,) + target_videos[0].shape).permute(3,0,4,1,5,2,6)
+    # target_video = target_video.reshape((-1, dur_t,) + size) ## Splicing Fragments
+    return target_video
+
+
+class FragmentSampleFrames:
+    def __init__(self, fsize_t, fragments_t, frame_interval=1, num_clips=1):
+
+        self.fragments_t = fragments_t
+        self.fsize_t = fsize_t
+        self.size_t = fragments_t * fsize_t
+        self.frame_interval = frame_interval
+        self.num_clips = num_clips
+
+    def get_frame_indices(self, num_frames):
+
+        tgrids = np.array(
+            [num_frames // self.fragments_t * i for i in range(self.fragments_t)],
+            dtype=np.int32,
+        )
+        tlength = num_frames // self.fragments_t
+
+        if tlength > self.fsize_t * self.frame_interval:
+            rnd_t = np.random.randint(
+                0, tlength - self.fsize_t * self.frame_interval, size=len(tgrids)
+            )
+        else:
+            rnd_t = np.zeros(len(tgrids), dtype=np.int32)
+
+        ranges_t = (
+            np.arange(self.fsize_t)[None, :] * self.frame_interval
+            + rnd_t[:, None]
+            + tgrids[:, None]
+        )
+        return np.concatenate(ranges_t)
+
+    def __call__(self, total_frames, train=False, start_index=0):
+        frame_inds = []
+        for i in range(self.num_clips):
+            frame_inds += [self.get_frame_indices(total_frames)]
+        frame_inds = np.concatenate(frame_inds)
+        frame_inds = np.mod(frame_inds + start_index, total_frames)
+        return frame_inds
+
+
+class SampleFrames:
+    def __init__(self, clip_len, frame_interval=1, num_clips=1):
+
+        self.clip_len = clip_len
+        self.frame_interval = frame_interval
+        self.num_clips = num_clips
+
+    def _get_train_clips(self, num_frames):
+        """Get clip offsets in train mode.
+
+        It will calculate the average interval for selected frames,
+        and randomly shift them within offsets between [0, avg_interval].
+        If the total number of frames is smaller than clips num or origin
+        frames length, it will return all zero indices.
+
+        Args:
+            num_frames (int): Total number of frame in the video.
+
+        Returns:
+            np.ndarray: Sampled frame indices in train mode.
+        """
+        ori_clip_len = self.clip_len * self.frame_interval
+        avg_interval = (num_frames - ori_clip_len + 1) // self.num_clips
+
+        if avg_interval > 0:
+            base_offsets = np.arange(self.num_clips) * avg_interval
+            clip_offsets = base_offsets + np.random.randint(
+                avg_interval, size=self.num_clips
+            )
+        elif num_frames > max(self.num_clips, ori_clip_len):
+            clip_offsets = np.sort(
+                np.random.randint(num_frames - ori_clip_len + 1, size=self.num_clips)
+            )
+        elif avg_interval == 0:
+            ratio = (num_frames - ori_clip_len + 1.0) / self.num_clips
+            clip_offsets = np.around(np.arange(self.num_clips) * ratio)
+        else:
+            clip_offsets = np.zeros((self.num_clips,), dtype=np.int)
+        return clip_offsets
+
+    def _get_test_clips(self, num_frames, start_index=0):
+        """Get clip offsets in test mode.
+
+        Calculate the average interval for selected frames, and shift them
+        fixedly by avg_interval/2.
+
+        Args:
+            num_frames (int): Total number of frame in the video.
+
+        Returns:
+            np.ndarray: Sampled frame indices in test mode.
+        """
+        ori_clip_len = self.clip_len * self.frame_interval
+        avg_interval = (num_frames - ori_clip_len + 1) / float(self.num_clips)
+        if num_frames > ori_clip_len - 1:
+            base_offsets = np.arange(self.num_clips) * avg_interval
+            clip_offsets = (base_offsets + avg_interval / 2.0).astype(np.int32)
+        else:
+            clip_offsets = np.zeros((self.num_clips,), dtype=np.int32)
+        return clip_offsets
+
+    def __call__(self, total_frames, train=False, start_index=0):
+        """Perform the SampleFrames loading.
+
+        Args:
+            results (dict): The resulting dict to be modified and passed
+                to the next transform in pipeline.
+        """
+        if train:
+            clip_offsets = self._get_train_clips(total_frames)
+        else:
+            clip_offsets = self._get_test_clips(total_frames)
+        frame_inds = (
+            clip_offsets[:, None]
+            + np.arange(self.clip_len)[None, :] * self.frame_interval
+        )
+        frame_inds = np.concatenate(frame_inds)
+
+        frame_inds = frame_inds.reshape((-1, self.clip_len))
+        frame_inds = np.mod(frame_inds, total_frames)
+        frame_inds = np.concatenate(frame_inds) + start_index
+        return frame_inds.astype(np.int32)
+
+
+class FastVQAPlusPlusDataset(torch.utils.data.Dataset):
+    def __init__(
+        self,
+        ann_file,
+        data_prefix,
+        frame_interval=2,
+        aligned=32,
+        fragments=(8, 8, 8),
+        fsize=(4, 32, 32),
+        num_clips=1,
+        nfrags=1,
+        cache_in_memory=False,
+        phase="test",
+        fallback_type="oversample",
+    ):
+        """
+        Fragments.
+        args:
+            fragments: G_f as in the paper.
+            fsize: S_f as in the paper.
+            nfrags: number of samples (spatially) as in the paper.
+            num_clips: number of samples (temporally) as in the paper.
+        """
+        self.ann_file = ann_file
+        self.data_prefix = data_prefix
+        self.frame_interval = frame_interval
+        self.num_clips = num_clips
+        self.fragments = fragments
+        self.fsize = fsize
+        self.nfrags = nfrags
+        self.clip_len = fragments[0] * fsize[0]
+        self.aligned = aligned
+        self.fallback_type = fallback_type
+        self.sampler = FragmentSampleFrames(
+            fsize[0], fragments[0], frame_interval, num_clips
+        )
+        self.video_infos = []
+        self.phase = phase
+        self.mean = torch.FloatTensor([123.675, 116.28, 103.53])
+        self.std = torch.FloatTensor([58.395, 57.12, 57.375])
+        if isinstance(self.ann_file, list):
+            self.video_infos = self.ann_file
+        else:
+            with open(self.ann_file, "r") as fin:
+                for line in fin:
+                    line_split = line.strip().split(",")
+                    filename, _, _, label = line_split
+                    label = float(label)
+                    filename = osp.join(self.data_prefix, filename)
+                    self.video_infos.append(dict(filename=filename, label=label))
+        if cache_in_memory:
+            self.cache = {}
+            for i in tqdm(range(len(self)), desc="Caching fragments"):
+                self.cache[i] = self.__getitem__(i, tocache=True)
+        else:
+            self.cache = None
+
+    def __getitem__(
+        self, index, tocache=False, need_original_frames=False,
+    ):
+        if tocache or self.cache is None:
+            fx, fy = self.fragments[1:]
+            fsx, fsy = self.fsize[1:]
+            video_info = self.video_infos[index]
+            filename = video_info["filename"]
+            label = video_info["label"]
+            if filename.endswith(".yuv"):
+                video = skvideo.io.vread(
+                    filename, 1080, 1920, inputdict={"-pix_fmt": "yuvj420p"}
+                )
+                frame_inds = self.sampler(video.shape[0], self.phase == "train")
+                imgs = [torch.from_numpy(video[idx]) for idx in frame_inds]
+            else:
+                vreader = VideoReader(filename)
+                frame_inds = self.sampler(len(vreader), self.phase == "train")
+                frame_dict = {idx: vreader[idx] for idx in np.unique(frame_inds)}
+                imgs = [frame_dict[idx] for idx in frame_inds]
+            img_shape = imgs[0].shape
+            video = torch.stack(imgs, 0)
+            video = video.permute(3, 0, 1, 2)
+            if self.nfrags == 1:
+                vfrag = get_spatial_fragments(
+                    video,
+                    fx,
+                    fy,
+                    fsx,
+                    fsy,
+                    aligned=self.aligned,
+                    fallback_type=self.fallback_type,
+                )
+            else:
+                vfrag = get_spatial_fragments(
+                    video,
+                    fx,
+                    fy,
+                    fsx,
+                    fsy,
+                    aligned=self.aligned,
+                    fallback_type=self.fallback_type,
+                )
+                for i in range(1, self.nfrags):
+                    vfrag = torch.cat(
+                        (
+                            vfrag,
+                            get_spatial_fragments(
+                                video,
+                                fragments,
+                                fx,
+                                fy,
+                                fsx,
+                                fsy,
+                                aligned=self.aligned,
+                                fallback_type=self.fallback_type,
+                            ),
+                        ),
+                        1,
+                    )
+            if tocache:
+                return (vfrag, frame_inds, label, img_shape)
+        else:
+            vfrag, frame_inds, label, img_shape = self.cache[index]
+        vfrag = ((vfrag.permute(1, 2, 3, 0) - self.mean) / self.std).permute(3, 0, 1, 2)
+        data = {
+            "video": vfrag.reshape(
+                (-1, self.nfrags * self.num_clips, self.clip_len) + vfrag.shape[2:]
+            ).transpose(
+                0, 1
+            ),  # B, V, T, C, H, W
+            "frame_inds": frame_inds,
+            "gt_label": label,
+            "original_shape": img_shape,
+        }
+        if need_original_frames:
+            data["original_video"] = video.reshape(
+                (-1, self.nfrags * self.num_clips, self.clip_len) + video.shape[2:]
+            ).transpose(0, 1)
+        return data
+
+    def __len__(self):
+        return len(self.video_infos)
+
+
+class FragmentVideoDataset(torch.utils.data.Dataset):
+    def __init__(
+        self,
+        ann_file,
+        data_prefix,
+        clip_len=32,
+        frame_interval=2,
+        num_clips=4,
+        aligned=32,
+        fragments=7,
+        fsize=32,
+        nfrags=1,
+        cache_in_memory=False,
+        phase="test",
+    ):
+        """
+        Fragments.
+        args:
+            fragments: G_f as in the paper.
+            fsize: S_f as in the paper.
+            nfrags: number of samples as in the paper.
+        """
+        self.ann_file = ann_file
+        self.data_prefix = data_prefix
+        self.clip_len = clip_len
+        self.frame_interval = frame_interval
+        self.num_clips = num_clips
+        self.fragments = fragments
+        self.fsize = fsize
+        self.nfrags = nfrags
+        self.aligned = aligned
+        self.sampler = SampleFrames(clip_len, frame_interval, num_clips)
+        self.video_infos = []
+        self.phase = phase
+        self.mean = torch.FloatTensor([123.675, 116.28, 103.53])
+        self.std = torch.FloatTensor([58.395, 57.12, 57.375])
+        if isinstance(self.ann_file, list):
+            self.video_infos = self.ann_file
+        else:
+            with open(self.ann_file, "r") as fin:
+                for line in fin:
+                    line_split = line.strip().split(",")
+                    filename, _, _, label = line_split
+                    label = float(label)
+                    filename = osp.join(self.data_prefix, filename)
+                    self.video_infos.append(dict(filename=filename, label=label))
+        if cache_in_memory:
+            self.cache = {}
+            for i in tqdm(range(len(self)), desc="Caching fragments"):
+                self.cache[i] = self.__getitem__(i, tocache=True)
+        else:
+            self.cache = None
+
+    def __getitem__(
+        self, index, fragments=-1, fsize=-1, tocache=False, need_original_frames=False,
+    ):
+        if tocache or self.cache is None:
+            if fragments == -1:
+                fragments = self.fragments
+            if fsize == -1:
+                fsize = self.fsize
+            video_info = self.video_infos[index]
+            filename = video_info["filename"]
+            label = video_info["label"]
+            if filename.endswith(".yuv"):
+                video = skvideo.io.vread(
+                    filename, 1080, 1920, inputdict={"-pix_fmt": "yuvj420p"}
+                )
+                frame_inds = self.sampler(video.shape[0], self.phase == "train")
+                imgs = [torch.from_numpy(video[idx]) for idx in frame_inds]
+            else:
+                vreader = VideoReader(filename)
+                frame_inds = self.sampler(len(vreader), self.phase == "train")
+                frame_dict = {idx: vreader[idx] for idx in np.unique(frame_inds)}
+                imgs = [frame_dict[idx] for idx in frame_inds]
+            img_shape = imgs[0].shape
+            video = torch.stack(imgs, 0)
+            video = video.permute(3, 0, 1, 2)
+            if self.nfrags == 1:
+                vfrag = get_spatial_fragments(
+                    video, fragments, fragments, fsize, fsize, aligned=self.aligned
+                )
+            else:
+                vfrag = get_spatial_fragments(
+                    video, fragments, fragments, fsize, fsize, aligned=self.aligned
+                )
+                for i in range(1, self.nfrags):
+                    vfrag = torch.cat(
+                        (
+                            vfrag,
+                            get_spatial_fragments(
+                                video,
+                                fragments,
+                                fragments,
+                                fsize,
+                                fsize,
+                                aligned=self.aligned,
+                            ),
+                        ),
+                        1,
+                    )
+            if tocache:
+                return (vfrag, frame_inds, label, img_shape)
+        else:
+            vfrag, frame_inds, label, img_shape = self.cache[index]
+        vfrag = ((vfrag.permute(1, 2, 3, 0) - self.mean) / self.std).permute(3, 0, 1, 2)
+        data = {
+            "video": vfrag.reshape(
+                (-1, self.nfrags * self.num_clips, self.clip_len) + vfrag.shape[2:]
+            ).transpose(
+                0, 1
+            ),  # B, V, T, C, H, W
+            "frame_inds": frame_inds,
+            "gt_label": label,
+            "original_shape": img_shape,
+        }
+        if need_original_frames:
+            data["original_video"] = video.reshape(
+                (-1, self.nfrags * self.num_clips, self.clip_len) + video.shape[2:]
+            ).transpose(0, 1)
+        return data
+
+    def __len__(self):
+        return len(self.video_infos)
+
+
+class ResizedVideoDataset(torch.utils.data.Dataset):
+    def __init__(
+        self,
+        ann_file,
+        data_prefix,
+        clip_len=32,
+        frame_interval=2,
+        num_clips=4,
+        aligned=32,
+        size=224,
+        cache_in_memory=False,
+        phase="test",
+    ):
+        """
+        Using resizing.
+        """
+        self.ann_file = ann_file
+        self.data_prefix = data_prefix
+        self.clip_len = clip_len
+        self.frame_interval = frame_interval
+        self.num_clips = num_clips
+        self.size = size
+        self.aligned = aligned
+        self.sampler = SampleFrames(clip_len, frame_interval, num_clips)
+        self.video_infos = []
+        self.phase = phase
+        self.mean = torch.FloatTensor([123.675, 116.28, 103.53])
+        self.std = torch.FloatTensor([58.395, 57.12, 57.375])
+        if isinstance(self.ann_file, list):
+            self.video_infos = self.ann_file
+        else:
+            with open(self.ann_file, "r") as fin:
+                for line in fin:
+                    line_split = line.strip().split(",")
+                    filename, _, _, label = line_split
+                    label = float(label)
+                    filename = osp.join(self.data_prefix, filename)
+                    self.video_infos.append(dict(filename=filename, label=label))
+        if cache_in_memory:
+            self.cache = {}
+            for i in tqdm(range(len(self)), desc="Caching resized videos"):
+                self.cache[i] = self.__getitem__(i, tocache=True)
+        else:
+            self.cache = None
+
+    def __getitem__(self, index, tocache=False, need_original_frames=False):
+        if tocache or self.cache is None:
+            video_info = self.video_infos[index]
+            filename = video_info["filename"]
+            label = video_info["label"]
+            vreader = VideoReader(filename)
+            frame_inds = self.sampler(len(vreader), self.phase == "train")
+            frame_dict = {idx: vreader[idx] for idx in np.unique(frame_inds)}
+            imgs = [frame_dict[idx] for idx in frame_inds]
+            img_shape = imgs[0].shape
+            video = torch.stack(imgs, 0)
+            video = video.permute(3, 0, 1, 2)
+            video = torch.nn.functional.interpolate(video, size=(self.size, self.size))
+            if tocache:
+                return (vfrag, frame_inds, label, img_shape)
+        else:
+            vfrag, frame_inds, label, img_shape = self.cache[index]
+        vfrag = ((vfrag.permute(1, 2, 3, 0) - self.mean) / self.std).permute(3, 0, 1, 2)
+        data = {
+            "video": vfrag.reshape(
+                (-1, self.num_clips, self.clip_len) + vfrag.shape[2:]
+            ).transpose(
+                0, 1
+            ),  # B, V, T, C, H, W
+            "frame_inds": frame_inds,
+            "gt_label": label,
+            "original_shape": img_shape,
+        }
+        if need_original_frames:
+            data["original_video"] = video.reshape(
+                (-1, self.nfrags * self.num_clips, self.clip_len) + video.shape[2:]
+            ).transpose(0, 1)
+        return data
+
+    def __len__(self):
+        return len(self.video_infos)
+
+
+class CroppedVideoDataset(FragmentVideoDataset):
+    def __init__(
+        self,
+        ann_file,
+        data_prefix,
+        clip_len=32,
+        frame_interval=2,
+        num_clips=4,
+        aligned=32,
+        size=224,
+        ncrops=1,
+        cache_in_memory=False,
+        phase="test",
+    ):
+
+        """
+        Regard Cropping as a special case for Fragments in Grid 1*1.
+        """
+        super().__init__(
+            ann_file,
+            data_prefix,
+            clip_len=clip_len,
+            frame_interval=frame_interval,
+            num_clips=num_clips,
+            aligned=aligned,
+            fragments=1,
+            fsize=224,
+            nfrags=ncrops,
+            cache_in_memory=cache_in_memory,
+            phase=phase,
+        )
+
+
+class FragmentImageDataset(torch.utils.data.Dataset):
+    def __init__(
+        self,
+        ann_file,
+        data_prefix,
+        fragments=7,
+        fsize=32,
+        nfrags=1,
+        cache_in_memory=False,
+        phase="test",
+    ):
+        self.ann_file = ann_file
+        self.data_prefix = data_prefix
+        self.fragments = fragments
+        self.fsize = fsize
+        self.nfrags = nfrags
+        self.image_infos = []
+        self.phase = phase
+        self.mean = torch.FloatTensor([123.675, 116.28, 103.53])
+        self.std = torch.FloatTensor([58.395, 57.12, 57.375])
+        if isinstance(self.ann_file, list):
+            self.image_infos = self.ann_file
+        else:
+            with open(self.ann_file, "r") as fin:
+                for line in fin:
+                    line_split = line.strip().split(",")
+                    filename, _, _, label = line_split
+                    label = float(label)
+                    filename = osp.join(self.data_prefix, filename)
+                    self.image_infos.append(dict(filename=filename, label=label))
+        if cache_in_memory:
+            self.cache = {}
+            for i in tqdm(range(len(self)), desc="Caching fragments"):
+                self.cache[i] = self.__getitem__(i, tocache=True)
+        else:
+            self.cache = None
+
+    def __getitem__(
+        self, index, fragments=-1, fsize=-1, tocache=False, need_original_frames=False
+    ):
+        if tocache or self.cache is None:
+            if fragments == -1:
+                fragments = self.fragments
+            if fsize == -1:
+                fsize = self.fsize
+            image_info = self.image_infos[index]
+            filename = image_info["filename"]
+            label = image_info["label"]
+            try:
+                img = torchvision.io.read_image(filename)
+            except:
+                img = cv2.imread(filename)
+                img = torch.from_numpy(img[:, :, [2, 1, 0]]).permute(2, 0, 1)
+            img_shape = img.shape[1:]
+            image = img.unsqueeze(1)
+            if self.nfrags == 1:
+                ifrag = get_spatial_fragments(image, fragments, fragments, fsize, fsize)
+            else:
+                ifrag = get_spatial_fragments(image, fragments, fragments, fsize, fsize)
+                for i in range(1, self.nfrags):
+                    ifrag = torch.cat(
+                        (
+                            ifrag,
+                            get_spatial_fragments(
+                                image, fragments, fragments, fsize, fsize
+                            ),
+                        ),
+                        1,
+                    )
+            if tocache:
+                return (ifrag, label, img_shape)
+        else:
+            ifrag, label, img_shape = self.cache[index]
+        if self.nfrags == 1:
+            ifrag = (
+                ((ifrag.permute(1, 2, 3, 0) - self.mean) / self.std)
+                .squeeze(0)
+                .permute(2, 0, 1)
+            )
+        else:
+            ### During testing, one image as a batch
+            ifrag = (
+                ((ifrag.permute(1, 2, 3, 0) - self.mean) / self.std)
+                .squeeze(0)
+                .permute(0, 3, 1, 2)
+            )
+        data = {
+            "image": ifrag,
+            "gt_label": label,
+            "original_shape": img_shape,
+            "name": filename,
+        }
+        if need_original_frames:
+            data["original_image"] = image.squeeze(1)
+        return data
+
+    def __len__(self):
+        return len(self.image_infos)
+
+
+class ResizedImageDataset(torch.utils.data.Dataset):
+    def __init__(
+        self, ann_file, data_prefix, size=224, cache_in_memory=False, phase="test",
+    ):
+        self.ann_file = ann_file
+        self.data_prefix = data_prefix
+        self.size = size
+        self.image_infos = []
+        self.phase = phase
+        self.mean = torch.FloatTensor([123.675, 116.28, 103.53])
+        self.std = torch.FloatTensor([58.395, 57.12, 57.375])
+        if isinstance(self.ann_file, list):
+            self.image_infos = self.ann_file
+        else:
+            with open(self.ann_file, "r") as fin:
+                for line in fin:
+                    line_split = line.strip().split(",")
+                    filename, _, _, label = line_split
+                    label = float(label)
+                    filename = osp.join(self.data_prefix, filename)
+                    self.image_infos.append(dict(filename=filename, label=label))
+        if cache_in_memory:
+            self.cache = {}
+            for i in tqdm(range(len(self)), desc="Caching fragments"):
+                self.cache[i] = self.__getitem__(i, tocache=True)
+        else:
+            self.cache = None
+
+    def __getitem__(
+        self, index, fragments=-1, fsize=-1, tocache=False, need_original_frames=False
+    ):
+        if tocache or self.cache is None:
+            if fragments == -1:
+                fragments = self.fragments
+            if fsize == -1:
+                fsize = self.fsize
+            image_info = self.image_infos[index]
+            filename = image_info["filename"]
+            label = image_info["label"]
+            img = torchvision.io.read_image(filename)
+            img_shape = img.shape[1:]
+            image = img.unsqueeze(1)
+            if self.nfrags == 1:
+                ifrag = get_spatial_fragments(image, fragments, fsize)
+            else:
+                ifrag = get_spatial_fragments(image, fragments, fsize)
+                for i in range(1, self.nfrags):
+                    ifrag = torch.cat(
+                        (ifrag, get_spatial_fragments(image, fragments, fsize)), 1
+                    )
+            if tocache:
+                return (ifrag, label, img_shape)
+        else:
+            ifrag, label, img_shape = self.cache[index]
+        ifrag = (
+            ((ifrag.permute(1, 2, 3, 0) - self.mean) / self.std)
+            .squeeze(0)
+            .permute(2, 0, 1)
+        )
+        data = {
+            "image": ifrag,
+            "gt_label": label,
+            "original_shape": img_shape,
+        }
+        if need_original_frames:
+            data["original_image"] = image.squeeze(1)
+        return data
+
+    def __len__(self):
+        return len(self.image_infos)
+
+
+class CroppedImageDataset(FragmentImageDataset):
+    def __init__(
+        self,
+        ann_file,
+        data_prefix,
+        size=224,
+        ncrops=1,
+        cache_in_memory=False,
+        phase="test",
+    ):
+
+        """
+        Regard Cropping as a special case for Fragments in Grid 1*1.
+        """
+        super().__init__(
+            ann_file,
+            data_prefix,
+            fragments=1,
+            fsize=224,
+            nfrags=ncrops,
+            cache_in_memory=cache_in_memory,
+            phase=phase,
+        )

cover/datasets/cover_datasets.py

@@ -0,0 +1,442 @@
+import copy
+import glob
+import os
+import os.path as osp
+import random
+from functools import lru_cache
+
+import cv2
+import decord
+import numpy as np
+import skvideo.io
+import torch
+import torchvision
+from decord import VideoReader, cpu, gpu
+from tqdm import tqdm
+
+random.seed(42)
+
+decord.bridge.set_bridge("torch")
+
+
+def get_spatial_fragments(
+    video,
+    fragments_h=7,
+    fragments_w=7,
+    fsize_h=32,
+    fsize_w=32,
+    aligned=32,
+    nfrags=1,
+    random=False,
+    random_upsample=False,
+    fallback_type="upsample",
+    upsample=-1,
+    **kwargs,
+):
+    if upsample > 0:
+        old_h, old_w = video.shape[-2], video.shape[-1]
+        if old_h >= old_w:
+            w = upsample
+            h = int(upsample * old_h / old_w)
+        else:
+            h = upsample
+            w = int(upsample * old_w / old_h)
+        
+        video = get_resized_video(video, h, w)
+    size_h = fragments_h * fsize_h
+    size_w = fragments_w * fsize_w
+    ## video: [C,T,H,W]
+    ## situation for images
+    if video.shape[1] == 1:
+        aligned = 1
+
+    dur_t, res_h, res_w = video.shape[-3:]
+    ratio = min(res_h / size_h, res_w / size_w)
+    if fallback_type == "upsample" and ratio < 1:
+
+        ovideo = video
+        video = torch.nn.functional.interpolate(
+            video / 255.0, scale_factor=1 / ratio, mode="bilinear"
+        )
+        video = (video * 255.0).type_as(ovideo)
+
+    if random_upsample:
+
+        randratio = random.random() * 0.5 + 1
+        video = torch.nn.functional.interpolate(
+            video / 255.0, scale_factor=randratio, mode="bilinear"
+        )
+        video = (video * 255.0).type_as(ovideo)
+        
+    assert dur_t % aligned == 0, "Please provide match vclip and align index"
+    size = size_h, size_w
+
+    ## make sure that sampling will not run out of the picture
+    hgrids = torch.LongTensor(
+        [min(res_h // fragments_h * i, res_h - fsize_h) for i in range(fragments_h)]
+    )
+    wgrids = torch.LongTensor(
+        [min(res_w // fragments_w * i, res_w - fsize_w) for i in range(fragments_w)]
+    )
+    hlength, wlength = res_h // fragments_h, res_w // fragments_w
+
+    if random:
+        print("This part is deprecated. Please remind that.")
+        if res_h > fsize_h:
+            rnd_h = torch.randint(
+                res_h - fsize_h, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_h = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+        if res_w > fsize_w:
+            rnd_w = torch.randint(
+                res_w - fsize_w, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_w = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+    else:
+        if hlength > fsize_h:
+            rnd_h = torch.randint(
+                hlength - fsize_h, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_h = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+        if wlength > fsize_w:
+            rnd_w = torch.randint(
+                wlength - fsize_w, (len(hgrids), len(wgrids), dur_t // aligned)
+            )
+        else:
+            rnd_w = torch.zeros((len(hgrids), len(wgrids), dur_t // aligned)).int()
+
+    target_video = torch.zeros(video.shape[:-2] + size).to(video.device)
+    # target_videos = []
+
+    for i, hs in enumerate(hgrids):
+        for j, ws in enumerate(wgrids):
+            for t in range(dur_t // aligned):
+                t_s, t_e = t * aligned, (t + 1) * aligned
+                h_s, h_e = i * fsize_h, (i + 1) * fsize_h
+                w_s, w_e = j * fsize_w, (j + 1) * fsize_w
+                if random:
+                    h_so, h_eo = rnd_h[i][j][t], rnd_h[i][j][t] + fsize_h
+                    w_so, w_eo = rnd_w[i][j][t], rnd_w[i][j][t] + fsize_w
+                else:
+                    h_so, h_eo = hs + rnd_h[i][j][t], hs + rnd_h[i][j][t] + fsize_h
+                    w_so, w_eo = ws + rnd_w[i][j][t], ws + rnd_w[i][j][t] + fsize_w
+                target_video[:, t_s:t_e, h_s:h_e, w_s:w_e] = video[
+                    :, t_s:t_e, h_so:h_eo, w_so:w_eo
+                ]
+    # target_videos.append(video[:,t_s:t_e,h_so:h_eo,w_so:w_eo])
+    # target_video = torch.stack(target_videos, 0).reshape((dur_t // aligned, fragments, fragments,) + target_videos[0].shape).permute(3,0,4,1,5,2,6)
+    # target_video = target_video.reshape((-1, dur_t,) + size) ## Splicing Fragments
+    return target_video
+
+
+@lru_cache
+def get_resize_function(size_h, size_w, target_ratio=1, random_crop=False):
+    if random_crop:
+        return torchvision.transforms.RandomResizedCrop(
+            (size_h, size_w), scale=(0.40, 1.0)
+        )
+    if target_ratio > 1:
+        size_h = int(target_ratio * size_w)
+        assert size_h > size_w
+    elif target_ratio < 1:
+        size_w = int(size_h / target_ratio)
+        assert size_w > size_h
+    return torchvision.transforms.Resize((size_h, size_w))
+
+
+def get_resized_video(
+    video, size_h=224, size_w=224, random_crop=False, arp=False, **kwargs,
+):
+    video = video.permute(1, 0, 2, 3)
+    resize_opt = get_resize_function(
+        size_h, size_w, video.shape[-2] / video.shape[-1] if arp else 1, random_crop
+    )
+    video = resize_opt(video).permute(1, 0, 2, 3)
+    return video
+
+
+def get_arp_resized_video(
+    video, short_edge=224, train=False, **kwargs,
+):
+    if train:  ## if during training, will random crop into square and then resize
+        res_h, res_w = video.shape[-2:]
+        ori_short_edge = min(video.shape[-2:])
+        if res_h > ori_short_edge:
+            rnd_h = random.randrange(res_h - ori_short_edge)
+            video = video[..., rnd_h : rnd_h + ori_short_edge, :]
+        elif res_w > ori_short_edge:
+            rnd_w = random.randrange(res_w - ori_short_edge)
+            video = video[..., :, rnd_h : rnd_h + ori_short_edge]
+    ori_short_edge = min(video.shape[-2:])
+    scale_factor = short_edge / ori_short_edge
+    ovideo = video
+    video = torch.nn.functional.interpolate(
+        video / 255.0, scale_factors=scale_factor, mode="bilinear"
+    )
+    video = (video * 255.0).type_as(ovideo)
+    return video
+
+
+def get_arp_fragment_video(
+    video, short_fragments=7, fsize=32, train=False, **kwargs,
+):
+    if (
+        train
+    ):  ## if during training, will random crop into square and then get fragments
+        res_h, res_w = video.shape[-2:]
+        ori_short_edge = min(video.shape[-2:])
+        if res_h > ori_short_edge:
+            rnd_h = random.randrange(res_h - ori_short_edge)
+            video = video[..., rnd_h : rnd_h + ori_short_edge, :]
+        elif res_w > ori_short_edge:
+            rnd_w = random.randrange(res_w - ori_short_edge)
+            video = video[..., :, rnd_h : rnd_h + ori_short_edge]
+    kwargs["fsize_h"], kwargs["fsize_w"] = fsize, fsize
+    res_h, res_w = video.shape[-2:]
+    if res_h > res_w:
+        kwargs["fragments_w"] = short_fragments
+        kwargs["fragments_h"] = int(short_fragments * res_h / res_w)
+    else:
+        kwargs["fragments_h"] = short_fragments
+        kwargs["fragments_w"] = int(short_fragments * res_w / res_h)
+    return get_spatial_fragments(video, **kwargs)
+
+
+def get_cropped_video(
+    video, size_h=224, size_w=224, **kwargs,
+):
+    kwargs["fragments_h"], kwargs["fragments_w"] = 1, 1
+    kwargs["fsize_h"], kwargs["fsize_w"] = size_h, size_w
+    return get_spatial_fragments(video, **kwargs)
+
+
+def get_single_view(
+    video, sample_type="aesthetic", **kwargs,
+):
+    if sample_type.startswith("aesthetic"):
+        video = get_resized_video(video, **kwargs)
+    elif sample_type.startswith("technical"):
+        video = get_spatial_fragments(video, **kwargs)
+    elif sample_type.startswith("semantic"):
+        video = get_resized_video(video, **kwargs)
+    elif sample_type == "original":
+        return video
+
+    return video
+
+
+def spatial_temporal_view_decomposition(
+    video_path, sample_types, samplers, is_train=False, augment=False,
+):
+    video = {}
+    if video_path.endswith(".yuv"):
+        print("This part will be deprecated due to large memory cost.")
+        ## This is only an adaptation to LIVE-Qualcomm
+        ovideo = skvideo.io.vread(
+            video_path, 1080, 1920, inputdict={"-pix_fmt": "yuvj420p"}
+        )
+        for stype in samplers:
+            frame_inds = samplers[stype](ovideo.shape[0], is_train)
+            imgs = [torch.from_numpy(ovideo[idx]) for idx in frame_inds]
+            video[stype] = torch.stack(imgs, 0).permute(3, 0, 1, 2)
+        del ovideo
+    else:
+        decord.bridge.set_bridge("torch")
+        vreader = VideoReader(video_path)
+        ### Avoid duplicated video decoding!!! Important!!!!
+        all_frame_inds = []
+        frame_inds = {}
+        for stype in samplers:
+            frame_inds[stype] = samplers[stype](len(vreader), is_train)
+            all_frame_inds.append(frame_inds[stype])
+
+        ### Each frame is only decoded one time!!!
+        all_frame_inds = np.concatenate(all_frame_inds, 0)
+        frame_dict = {idx: vreader[idx] for idx in np.unique(all_frame_inds)}
+
+        for stype in samplers:
+            imgs = [frame_dict[idx] for idx in frame_inds[stype]]
+            video[stype] = torch.stack(imgs, 0).permute(3, 0, 1, 2)
+
+    sampled_video = {}
+    for stype, sopt in sample_types.items():
+        sampled_video[stype] = get_single_view(video[stype], stype, **sopt)
+    return sampled_video, frame_inds
+
+
+import random
+
+import numpy as np
+
+
+class UnifiedFrameSampler:
+    def __init__(
+        self, fsize_t, fragments_t, frame_interval=1, num_clips=1, drop_rate=0.0,
+    ):
+
+        self.fragments_t = fragments_t
+        self.fsize_t = fsize_t
+        self.size_t = fragments_t * fsize_t
+        self.frame_interval = frame_interval
+        self.num_clips = num_clips
+        self.drop_rate = drop_rate
+
+    def get_frame_indices(self, num_frames, train=False):
+
+        tgrids = np.array(
+            [num_frames // self.fragments_t * i for i in range(self.fragments_t)],
+            dtype=np.int32,
+        )
+        tlength = num_frames // self.fragments_t
+
+        if tlength > self.fsize_t * self.frame_interval:
+            rnd_t = np.random.randint(
+                0, tlength - self.fsize_t * self.frame_interval, size=len(tgrids)
+            )
+        else:
+            rnd_t = np.zeros(len(tgrids), dtype=np.int32)
+
+        ranges_t = (
+            np.arange(self.fsize_t)[None, :] * self.frame_interval
+            + rnd_t[:, None]
+            + tgrids[:, None]
+        )
+
+        drop = random.sample(
+            list(range(self.fragments_t)), int(self.fragments_t * self.drop_rate)
+        )
+        dropped_ranges_t = []
+        for i, rt in enumerate(ranges_t):
+            if i not in drop:
+                dropped_ranges_t.append(rt)
+        return np.concatenate(dropped_ranges_t)
+
+    def __call__(self, total_frames, train=False, start_index=0):
+        frame_inds = []
+
+        for i in range(self.num_clips):
+            frame_inds += [self.get_frame_indices(total_frames)]
+
+        frame_inds = np.concatenate(frame_inds)
+        frame_inds = np.mod(frame_inds + start_index, total_frames)
+        return frame_inds.astype(np.int32)
+
+
+class ViewDecompositionDataset(torch.utils.data.Dataset):
+    def __init__(self, opt):
+        ## opt is a dictionary that includes options for video sampling
+
+        super().__init__()
+
+        self.weight = opt.get("weight", 0.5)
+        
+        self.fully_supervised = opt.get("fully_supervised", False)
+        print("Fully supervised:", self.fully_supervised)
+        
+        self.video_infos = []
+        self.ann_file = opt["anno_file"]
+        self.data_prefix = opt["data_prefix"]
+        self.opt = opt
+        self.sample_types = opt["sample_types"]
+        self.data_backend = opt.get("data_backend", "disk")
+        self.augment = opt.get("augment", False)
+        if self.data_backend == "petrel":
+            from petrel_client import client
+
+            self.client = client.Client(enable_mc=True)
+
+        self.phase = opt["phase"]
+        self.crop = opt.get("random_crop", False)
+        self.mean = torch.FloatTensor([123.675, 116.28, 103.53])
+        self.std = torch.FloatTensor([58.395, 57.12, 57.375])
+        self.mean_semantic = torch.FloatTensor([122.77, 116.75, 104.09])
+        self.std_semantic = torch.FloatTensor([68.50, 66.63, 70.32])
+        self.samplers = {}
+        for stype, sopt in opt["sample_types"].items():
+            if "t_frag" not in sopt:
+                # resized temporal sampling for TQE in COVER
+                self.samplers[stype] = UnifiedFrameSampler(
+                    sopt["clip_len"], sopt["num_clips"], sopt["frame_interval"]
+                )
+            else:
+                # temporal sampling for AQE in COVER
+                self.samplers[stype] = UnifiedFrameSampler(
+                    sopt["clip_len"] // sopt["t_frag"],
+                    sopt["t_frag"],
+                    sopt["frame_interval"],
+                    sopt["num_clips"],
+                )
+            print(
+                stype + " branch sampled frames:",
+                self.samplers[stype](240, self.phase == "train"),
+            )
+
+        if isinstance(self.ann_file, list):
+            self.video_infos = self.ann_file
+        else:
+            try:
+                with open(self.ann_file, "r") as fin:
+                    for line in fin:
+                        line_split = line.strip().split(",")
+                        filename, a, t, label = line_split
+                        if self.fully_supervised:
+                            label = float(a), float(t), float(label)
+                        else:
+                            label = float(label)
+                        filename = osp.join(self.data_prefix, filename)
+                        self.video_infos.append(dict(filename=filename, label=label))
+            except:
+                #### No Label Testing
+                video_filenames = []
+                for (root, dirs, files) in os.walk(self.data_prefix, topdown=True):
+                    for file in files:
+                        if file.endswith(".mp4"):
+                            video_filenames += [os.path.join(root, file)]
+                print(len(video_filenames))
+                video_filenames = sorted(video_filenames)
+                for filename in video_filenames:
+                    self.video_infos.append(dict(filename=filename, label=-1))
+
+    def __getitem__(self, index):
+        video_info = self.video_infos[index]
+        filename = video_info["filename"]
+        label = video_info["label"]
+
+        try:
+            ## Read Original Frames
+            ## Process Frames
+            data, frame_inds = spatial_temporal_view_decomposition(
+                filename,
+                self.sample_types,
+                self.samplers,
+                self.phase == "train",
+                self.augment and (self.phase == "train"),
+            )
+
+            for k, v in data.items():
+                if k == 'technical' or k == 'aesthetic':
+                    data[k] = ((v.permute(1, 2, 3, 0) - self.mean) / self.std).permute(
+                        3, 0, 1, 2
+                    )
+                elif k == 'semantic' :
+                    data[k] = ((v.permute(1, 2, 3, 0) - self.mean_semantic) / self.std_semantic).permute(
+                        3, 0, 1, 2
+                    )
+
+            data["num_clips"] = {}
+            for stype, sopt in self.sample_types.items():
+                data["num_clips"][stype] = sopt["num_clips"]
+            data["frame_inds"] = frame_inds
+            data["gt_label"] = label
+            data["name"] = filename  # osp.basename(video_info["filename"])
+        except:
+            # exception flow
+            return {"name": filename}
+
+        return data
+
+    def __len__(self):
+        return len(self.video_infos)

cover/models/__init__.py

@@ -0,0 +1,17 @@
+from .conv_backbone import convnext_3d_small, convnext_3d_tiny
+from .evaluator import COVER, BaseEvaluator, BaseImageEvaluator
+from .head import IQAHead, VARHead, VQAHead
+from .swin_backbone import SwinTransformer2D as IQABackbone
+from .swin_backbone import SwinTransformer3D as VQABackbone
+from .swin_backbone import swin_3d_small, swin_3d_tiny
+
+__all__ = [
+    "VQABackbone",
+    "IQABackbone",
+    "VQAHead",
+    "IQAHead",
+    "VARHead",
+    "BaseEvaluator",
+    "BaseImageEvaluator",
+    "COVER",
+]

cover/models/backbone_get_attention.py

@@ -0,0 +1,990 @@
+from functools import lru_cache, reduce
+from operator import mul
+
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.utils.checkpoint as checkpoint
+from einops import rearrange
+from timm.models.layers import DropPath, trunc_normal_
+
+
+def fragment_infos(D, H, W, fragments=7, device="cuda"):
+    m = torch.arange(fragments).unsqueeze(-1).float()
+    m = (m + m.t() * fragments).reshape(1, 1, 1, fragments, fragments)
+    m = F.interpolate(m.to(device), size=(D, H, W)).permute(0, 2, 3, 4, 1)
+    return m.long()
+
+
+@lru_cache
+def global_position_index(
+    D,
+    H,
+    W,
+    fragments=(1, 7, 7),
+    window_size=(8, 7, 7),
+    shift_size=(0, 0, 0),
+    device="cuda",
+):
+    frags_d = torch.arange(fragments[0])
+    frags_h = torch.arange(fragments[1])
+    frags_w = torch.arange(fragments[2])
+    frags = torch.stack(
+        torch.meshgrid(frags_d, frags_h, frags_w)
+    ).float()  # 3, Fd, Fh, Fw
+    coords = (
+        torch.nn.functional.interpolate(frags[None].to(device), size=(D, H, W))
+        .long()
+        .permute(0, 2, 3, 4, 1)
+    )
+    # print(shift_size)
+    coords = torch.roll(
+        coords, shifts=(-shift_size[0], -shift_size[1], -shift_size[2]), dims=(1, 2, 3)
+    )
+    window_coords = window_partition(coords, window_size)
+    relative_coords = (
+        window_coords[:, None, :] - window_coords[:, :, None]
+    )  # Wd*Wh*Ww, Wd*Wh*Ww, 3
+    return relative_coords  # relative_coords
+
+
+class Mlp(nn.Module):
+    """Multilayer perceptron."""
+
+    def __init__(
+        self,
+        in_features,
+        hidden_features=None,
+        out_features=None,
+        act_layer=nn.GELU,
+        drop=0.0,
+    ):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.fc1 = nn.Linear(in_features, hidden_features)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features)
+        self.drop = nn.Dropout(drop)
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.drop(x)
+        x = self.fc2(x)
+        x = self.drop(x)
+        return x
+
+
+def window_partition(x, window_size):
+    """
+    Args:
+        x: (B, D, H, W, C)
+        window_size (tuple[int]): window size
+
+    Returns:
+        windows: (B*num_windows, window_size*window_size, C)
+    """
+    B, D, H, W, C = x.shape
+    x = x.view(
+        B,
+        D // window_size[0],
+        window_size[0],
+        H // window_size[1],
+        window_size[1],
+        W // window_size[2],
+        window_size[2],
+        C,
+    )
+    windows = (
+        x.permute(0, 1, 3, 5, 2, 4, 6, 7)
+        .contiguous()
+        .view(-1, reduce(mul, window_size), C)
+    )
+    return windows
+
+
+def window_reverse(windows, window_size, B, D, H, W):
+    """
+    Args:
+        windows: (B*num_windows, window_size, window_size, C)
+        window_size (tuple[int]): Window size
+        H (int): Height of image
+        W (int): Width of image
+
+    Returns:
+        x: (B, D, H, W, C)
+    """
+    x = windows.view(
+        B,
+        D // window_size[0],
+        H // window_size[1],
+        W // window_size[2],
+        window_size[0],
+        window_size[1],
+        window_size[2],
+        -1,
+    )
+    x = x.permute(0, 1, 4, 2, 5, 3, 6, 7).contiguous().view(B, D, H, W, -1)
+    return x
+
+
+def get_window_size(x_size, window_size, shift_size=None):
+    use_window_size = list(window_size)
+    if shift_size is not None:
+        use_shift_size = list(shift_size)
+    for i in range(len(x_size)):
+        if x_size[i] <= window_size[i]:
+            use_window_size[i] = x_size[i]
+            if shift_size is not None:
+                use_shift_size[i] = 0
+
+    if shift_size is None:
+        return tuple(use_window_size)
+    else:
+        return tuple(use_window_size), tuple(use_shift_size)
+
+
+class WindowAttention3D(nn.Module):
+    """Window based multi-head self attention (W-MSA) module with relative position bias.
+    It supports both of shifted and non-shifted window.
+    Args:
+        dim (int): Number of input channels.
+        window_size (tuple[int]): The temporal length, height and width of the window.
+        num_heads (int): Number of attention heads.
+        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set
+        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
+        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
+    """
+
+    def __init__(
+        self,
+        dim,
+        window_size,
+        num_heads,
+        qkv_bias=False,
+        qk_scale=None,
+        attn_drop=0.0,
+        proj_drop=0.0,
+        frag_bias=False,
+    ):
+
+        super().__init__()
+        self.dim = dim
+        self.window_size = window_size  # Wd, Wh, Ww
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.scale = qk_scale or head_dim ** -0.5
+
+        # define a parameter table of relative position bias
+        self.relative_position_bias_table = nn.Parameter(
+            torch.zeros(
+                (2 * window_size[0] - 1)
+                * (2 * window_size[1] - 1)
+                * (2 * window_size[2] - 1),
+                num_heads,
+            )
+        )  # 2*Wd-1 * 2*Wh-1 * 2*Ww-1, nH
+        if frag_bias:
+            self.fragment_position_bias_table = nn.Parameter(
+                torch.zeros(
+                    (2 * window_size[0] - 1)
+                    * (2 * window_size[1] - 1)
+                    * (2 * window_size[2] - 1),
+                    num_heads,
+                )
+            )
+
+        # get pair-wise relative position index for each token inside the window
+        coords_d = torch.arange(self.window_size[0])
+        coords_h = torch.arange(self.window_size[1])
+        coords_w = torch.arange(self.window_size[2])
+        coords = torch.stack(
+            torch.meshgrid(coords_d, coords_h, coords_w)
+        )  # 3, Wd, Wh, Ww
+        coords_flatten = torch.flatten(coords, 1)  # 3, Wd*Wh*Ww
+        relative_coords = (
+            coords_flatten[:, :, None] - coords_flatten[:, None, :]
+        )  # 3, Wd*Wh*Ww, Wd*Wh*Ww
+        relative_coords = relative_coords.permute(
+            1, 2, 0
+        ).contiguous()  # Wd*Wh*Ww, Wd*Wh*Ww, 3
+        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0
+        relative_coords[:, :, 1] += self.window_size[1] - 1
+        relative_coords[:, :, 2] += self.window_size[2] - 1
+
+        relative_coords[:, :, 0] *= (2 * self.window_size[1] - 1) * (
+            2 * self.window_size[2] - 1
+        )
+        relative_coords[:, :, 1] *= 2 * self.window_size[2] - 1
+        relative_position_index = relative_coords.sum(-1)  # Wd*Wh*Ww, Wd*Wh*Ww
+        self.register_buffer("relative_position_index", relative_position_index)
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(dim, dim)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+        trunc_normal_(self.relative_position_bias_table, std=0.02)
+        self.softmax = nn.Softmax(dim=-1)
+
+    def forward(self, x, mask=None, fmask=None):
+        """Forward function.
+        Args:
+            x: input features with shape of (num_windows*B, N, C)
+            mask: (0/-inf) mask with shape of (num_windows, N, N) or None
+        """
+        B_, N, C = x.shape
+        qkv = (
+            self.qkv(x)
+            .reshape(B_, N, 3, self.num_heads, C // self.num_heads)
+            .permute(2, 0, 3, 1, 4)
+        )
+        q, k, v = qkv[0], qkv[1], qkv[2]  # B_, nH, N, C
+
+        q = q * self.scale
+        attn = q @ k.transpose(-2, -1)
+
+        relative_position_bias = self.relative_position_bias_table[
+            self.relative_position_index[:N, :N].reshape(-1)
+        ].reshape(
+            N, N, -1
+        )  # Wd*Wh*Ww,Wd*Wh*Ww,nH
+        relative_position_bias = relative_position_bias.permute(
+            2, 0, 1
+        ).contiguous()  # nH, Wd*Wh*Ww, Wd*Wh*Ww
+        if hasattr(self, "fragment_position_bias_table"):
+            fragment_position_bias = self.fragment_position_bias_table[
+                self.relative_position_index[:N, :N].reshape(-1)
+            ].reshape(
+                N, N, -1
+            )  # Wd*Wh*Ww,Wd*Wh*Ww,nH
+            fragment_position_bias = fragment_position_bias.permute(
+                2, 0, 1
+            ).contiguous()  # nH, Wd*Wh*Ww, Wd*Wh*Ww
+
+        ### Mask Position Bias
+        if fmask is not None:
+            # fgate = torch.where(fmask - fmask.transpose(-1, -2) == 0, 1, 0).float()
+            fgate = fmask.abs().sum(-1)
+            nW = fmask.shape[0]
+            relative_position_bias = relative_position_bias.unsqueeze(0)
+            fgate = fgate.unsqueeze(1)
+            # print(fgate.shape, relative_position_bias.shape)
+            if hasattr(self, "fragment_position_bias_table"):
+                relative_position_bias = (
+                    relative_position_bias * fgate
+                    + fragment_position_bias * (1 - fgate)
+                )
+
+            attn = attn.view(
+                B_ // nW, nW, self.num_heads, N, N
+            ) + relative_position_bias.unsqueeze(0)
+            attn = attn.view(-1, self.num_heads, N, N)
+        else:
+            attn = attn + relative_position_bias.unsqueeze(0)  # B_, nH, N, N
+
+        if mask is not None:
+            nW = mask.shape[0]
+            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(
+                1
+            ).unsqueeze(0)
+            attn = attn.view(-1, self.num_heads, N, N)
+            attn = self.softmax(attn)
+        else:
+            attn = self.softmax(attn)
+        attn = self.attn_drop(attn)
+
+        if B_ < 16:
+            avg_attn = (attn.mean((1, 2)).detach(), attn.mean((1, 3)).detach())
+        else:
+            avg_attn = None
+
+        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+
+        return x, avg_attn
+
+
+class SwinTransformerBlock3D(nn.Module):
+    """Swin Transformer Block.
+
+    Args:
+        dim (int): Number of input channels.
+        num_heads (int): Number of attention heads.
+        window_size (tuple[int]): Window size.
+        shift_size (tuple[int]): Shift size for SW-MSA.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float, optional): Stochastic depth rate. Default: 0.0
+        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+
+    def __init__(
+        self,
+        dim,
+        num_heads,
+        window_size=(2, 7, 7),
+        shift_size=(0, 0, 0),
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        qk_scale=None,
+        drop=0.0,
+        attn_drop=0.0,
+        drop_path=0.0,
+        act_layer=nn.GELU,
+        norm_layer=nn.LayerNorm,
+        use_checkpoint=False,
+        jump_attention=False,
+        frag_bias=False,
+    ):
+        super().__init__()
+        self.dim = dim
+        self.num_heads = num_heads
+        self.window_size = window_size
+        self.shift_size = shift_size
+        self.mlp_ratio = mlp_ratio
+        self.use_checkpoint = use_checkpoint
+        self.jump_attention = jump_attention
+        self.frag_bias = frag_bias
+
+        assert (
+            0 <= self.shift_size[0] < self.window_size[0]
+        ), "shift_size must in 0-window_size"
+        assert (
+            0 <= self.shift_size[1] < self.window_size[1]
+        ), "shift_size must in 0-window_size"
+        assert (
+            0 <= self.shift_size[2] < self.window_size[2]
+        ), "shift_size must in 0-window_size"
+
+        self.norm1 = norm_layer(dim)
+        self.attn = WindowAttention3D(
+            dim,
+            window_size=self.window_size,
+            num_heads=num_heads,
+            qkv_bias=qkv_bias,
+            qk_scale=qk_scale,
+            attn_drop=attn_drop,
+            proj_drop=drop,
+            frag_bias=frag_bias,
+        )
+
+        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
+        self.norm2 = norm_layer(dim)
+        mlp_hidden_dim = int(dim * mlp_ratio)
+        self.mlp = Mlp(
+            in_features=dim,
+            hidden_features=mlp_hidden_dim,
+            act_layer=act_layer,
+            drop=drop,
+        )
+
+    def forward_part1(self, x, mask_matrix):
+        B, D, H, W, C = x.shape
+        window_size, shift_size = get_window_size(
+            (D, H, W), self.window_size, self.shift_size
+        )
+
+        x = self.norm1(x)
+        # pad feature maps to multiples of window size
+        pad_l = pad_t = pad_d0 = 0
+        pad_d1 = (window_size[0] - D % window_size[0]) % window_size[0]
+        pad_b = (window_size[1] - H % window_size[1]) % window_size[1]
+        pad_r = (window_size[2] - W % window_size[2]) % window_size[2]
+
+        x = F.pad(x, (0, 0, pad_l, pad_r, pad_t, pad_b, pad_d0, pad_d1))
+        _, Dp, Hp, Wp, _ = x.shape
+        if False:  # not hasattr(self, 'finfo_windows'):
+            finfo = fragment_infos(Dp, Hp, Wp)
+
+        # cyclic shift
+        if any(i > 0 for i in shift_size):
+            shifted_x = torch.roll(
+                x,
+                shifts=(-shift_size[0], -shift_size[1], -shift_size[2]),
+                dims=(1, 2, 3),
+            )
+            if False:  # not hasattr(self, 'finfo_windows'):
+                shifted_finfo = torch.roll(
+                    finfo,
+                    shifts=(-shift_size[0], -shift_size[1], -shift_size[2]),
+                    dims=(1, 2, 3),
+                )
+            attn_mask = mask_matrix
+        else:
+            shifted_x = x
+            if False:  # not hasattr(self, 'finfo_windows'):
+                shifted_finfo = finfo
+            attn_mask = None
+        # partition windows
+        x_windows = window_partition(shifted_x, window_size)  # B*nW, Wd*Wh*Ww, C
+        if False:  # not hasattr(self, 'finfo_windows'):
+            self.finfo_windows = window_partition(shifted_finfo, window_size)
+        # W-MSA/SW-MSA
+        # print(shift_size)
+        gpi = global_position_index(
+            Dp, Hp, Wp, window_size=window_size, shift_size=shift_size, device=x.device
+        )
+        attn_windows, avg_attn = self.attn(
+            x_windows, mask=attn_mask, fmask=gpi
+        )  # self.finfo_windows)  # B*nW, Wd*Wh*Ww, C
+        # merge windows
+        attn_windows = attn_windows.view(-1, *(window_size + (C,)))
+        shifted_x = window_reverse(
+            attn_windows, window_size, B, Dp, Hp, Wp
+        )  # B D' H' W' C
+        # reverse cyclic shift
+        if any(i > 0 for i in shift_size):
+            x = torch.roll(
+                shifted_x,
+                shifts=(shift_size[0], shift_size[1], shift_size[2]),
+                dims=(1, 2, 3),
+            )
+        else:
+            x = shifted_x
+
+        if pad_d1 > 0 or pad_r > 0 or pad_b > 0:
+            x = x[:, :D, :H, :W, :].contiguous()
+        return x, avg_attn
+
+    def forward_part2(self, x):
+        return self.drop_path(self.mlp(self.norm2(x)))
+
+    def forward(self, x, mask_matrix):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, D, H, W, C).
+            mask_matrix: Attention mask for cyclic shift.
+        """
+
+        shortcut = x
+        if not self.jump_attention:
+            if self.use_checkpoint:
+                x = checkpoint.checkpoint(self.forward_part1, x, mask_matrix)
+            else:
+                x, avg_attn = self.forward_part1(x, mask_matrix)
+            x = shortcut + self.drop_path(x)
+
+        if self.use_checkpoint:
+            x = x + checkpoint.checkpoint(self.forward_part2, x)
+        else:
+            x = x + self.forward_part2(x)
+
+        return x, avg_attn
+
+
+class PatchMerging(nn.Module):
+    """Patch Merging Layer
+
+    Args:
+        dim (int): Number of input channels.
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+
+    def __init__(self, dim, norm_layer=nn.LayerNorm):
+        super().__init__()
+        self.dim = dim
+        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
+        self.norm = norm_layer(4 * dim)
+
+    def forward(self, x):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, D, H, W, C).
+        """
+        B, D, H, W, C = x.shape
+
+        # padding
+        pad_input = (H % 2 == 1) or (W % 2 == 1)
+        if pad_input:
+            x = F.pad(x, (0, 0, 0, W % 2, 0, H % 2))
+
+        x0 = x[:, :, 0::2, 0::2, :]  # B D H/2 W/2 C
+        x1 = x[:, :, 1::2, 0::2, :]  # B D H/2 W/2 C
+        x2 = x[:, :, 0::2, 1::2, :]  # B D H/2 W/2 C
+        x3 = x[:, :, 1::2, 1::2, :]  # B D H/2 W/2 C
+        x = torch.cat([x0, x1, x2, x3], -1)  # B D H/2 W/2 4*C
+
+        x = self.norm(x)
+        x = self.reduction(x)
+
+        return x
+
+
+# cache each stage results
+@lru_cache()
+def compute_mask(D, H, W, window_size, shift_size, device):
+    img_mask = torch.zeros((1, D, H, W, 1), device=device)  # 1 Dp Hp Wp 1
+    cnt = 0
+    for d in (
+        slice(-window_size[0]),
+        slice(-window_size[0], -shift_size[0]),
+        slice(-shift_size[0], None),
+    ):
+        for h in (
+            slice(-window_size[1]),
+            slice(-window_size[1], -shift_size[1]),
+            slice(-shift_size[1], None),
+        ):
+            for w in (
+                slice(-window_size[2]),
+                slice(-window_size[2], -shift_size[2]),
+                slice(-shift_size[2], None),
+            ):
+                img_mask[:, d, h, w, :] = cnt
+                cnt += 1
+    mask_windows = window_partition(img_mask, window_size)  # nW, ws[0]*ws[1]*ws[2], 1
+    mask_windows = mask_windows.squeeze(-1)  # nW, ws[0]*ws[1]*ws[2]
+    attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
+    attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(
+        attn_mask == 0, float(0.0)
+    )
+    return attn_mask
+
+
+class BasicLayer(nn.Module):
+    """A basic Swin Transformer layer for one stage.
+
+    Args:
+        dim (int): Number of feature channels
+        depth (int): Depths of this stage.
+        num_heads (int): Number of attention head.
+        window_size (tuple[int]): Local window size. Default: (1,7,7).
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
+        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
+        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
+    """
+
+    def __init__(
+        self,
+        dim,
+        depth,
+        num_heads,
+        window_size=(1, 7, 7),
+        mlp_ratio=4.0,
+        qkv_bias=False,
+        qk_scale=None,
+        drop=0.0,
+        attn_drop=0.0,
+        drop_path=0.0,
+        norm_layer=nn.LayerNorm,
+        downsample=None,
+        use_checkpoint=False,
+        jump_attention=False,
+        frag_bias=False,
+    ):
+        super().__init__()
+        self.window_size = window_size
+        self.shift_size = tuple(i // 2 for i in window_size)
+        self.depth = depth
+        self.use_checkpoint = use_checkpoint
+
+        # build blocks
+        self.blocks = nn.ModuleList(
+            [
+                SwinTransformerBlock3D(
+                    dim=dim,
+                    num_heads=num_heads,
+                    window_size=window_size,
+                    shift_size=(0, 0, 0) if (i % 2 == 0) else self.shift_size,
+                    mlp_ratio=mlp_ratio,
+                    qkv_bias=qkv_bias,
+                    qk_scale=qk_scale,
+                    drop=drop,
+                    attn_drop=attn_drop,
+                    drop_path=drop_path[i]
+                    if isinstance(drop_path, list)
+                    else drop_path,
+                    norm_layer=norm_layer,
+                    use_checkpoint=use_checkpoint,
+                    jump_attention=jump_attention,
+                    frag_bias=frag_bias,
+                )
+                for i in range(depth)
+            ]
+        )
+
+        self.downsample = downsample
+        if self.downsample is not None:
+            self.downsample = downsample(dim=dim, norm_layer=norm_layer)
+
+    def forward(self, x):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, C, D, H, W).
+        """
+        # calculate attention mask for SW-MSA
+        B, C, D, H, W = x.shape
+        window_size, shift_size = get_window_size(
+            (D, H, W), self.window_size, self.shift_size
+        )
+        x = rearrange(x, "b c d h w -> b d h w c")
+        Dp = int(np.ceil(D / window_size[0])) * window_size[0]
+        Hp = int(np.ceil(H / window_size[1])) * window_size[1]
+        Wp = int(np.ceil(W / window_size[2])) * window_size[2]
+        attn_mask = compute_mask(Dp, Hp, Wp, window_size, shift_size, x.device)
+        avg_attns = []
+        for blk in self.blocks:
+            x, avg_attn = blk(x, attn_mask)
+            if avg_attn is not None:
+                avg_attns.append(avg_attn)
+        x = x.view(B, D, H, W, -1)
+
+        if self.downsample is not None:
+            x = self.downsample(x)
+        x = rearrange(x, "b d h w c -> b c d h w")
+        return x, avg_attns
+
+
+class PatchEmbed3D(nn.Module):
+    """Video to Patch Embedding.
+
+    Args:
+        patch_size (int): Patch token size. Default: (2,4,4).
+        in_chans (int): Number of input video channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        norm_layer (nn.Module, optional): Normalization layer. Default: None
+    """
+
+    def __init__(self, patch_size=(2, 4, 4), in_chans=3, embed_dim=96, norm_layer=None):
+        super().__init__()
+        self.patch_size = patch_size
+
+        self.in_chans = in_chans
+        self.embed_dim = embed_dim
+
+        self.proj = nn.Conv3d(
+            in_chans, embed_dim, kernel_size=patch_size, stride=patch_size
+        )
+        if norm_layer is not None:
+            self.norm = norm_layer(embed_dim)
+        else:
+            self.norm = None
+
+    def forward(self, x):
+        """Forward function."""
+        # padding
+        _, _, D, H, W = x.size()
+        if W % self.patch_size[2] != 0:
+            x = F.pad(x, (0, self.patch_size[2] - W % self.patch_size[2]))
+        if H % self.patch_size[1] != 0:
+            x = F.pad(x, (0, 0, 0, self.patch_size[1] - H % self.patch_size[1]))
+        if D % self.patch_size[0] != 0:
+            x = F.pad(x, (0, 0, 0, 0, 0, self.patch_size[0] - D % self.patch_size[0]))
+
+        x = self.proj(x)  # B C D Wh Ww
+        if self.norm is not None:
+            D, Wh, Ww = x.size(2), x.size(3), x.size(4)
+            x = x.flatten(2).transpose(1, 2)
+            x = self.norm(x)
+            x = x.transpose(1, 2).view(-1, self.embed_dim, D, Wh, Ww)
+
+        return x
+
+
+class SwinTransformer3D(nn.Module):
+    """Swin Transformer backbone.
+        A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows`  -
+          https://arxiv.org/pdf/2103.14030
+
+    Args:
+        patch_size (int | tuple(int)): Patch size. Default: (4,4,4).
+        in_chans (int): Number of input image channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        depths (tuple[int]): Depths of each Swin Transformer stage.
+        num_heads (tuple[int]): Number of attention head of each stage.
+        window_size (int): Window size. Default: 7.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: Truee
+        qk_scale (float): Override default qk scale of head_dim ** -0.5 if set.
+        drop_rate (float): Dropout rate.
+        attn_drop_rate (float): Attention dropout rate. Default: 0.
+        drop_path_rate (float): Stochastic depth rate. Default: 0.2.
+        norm_layer: Normalization layer. Default: nn.LayerNorm.
+        patch_norm (bool): If True, add normalization after patch embedding. Default: False.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters.
+    """
+
+    def __init__(
+        self,
+        pretrained=None,
+        pretrained2d=False,
+        patch_size=(2, 4, 4),
+        in_chans=3,
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=(8, 7, 7),
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        qk_scale=None,
+        drop_rate=0.0,
+        attn_drop_rate=0.0,
+        drop_path_rate=0.1,
+        norm_layer=nn.LayerNorm,
+        patch_norm=True,
+        frozen_stages=-1,
+        use_checkpoint=True,
+        jump_attention=[False, False, False, False],
+        frag_biases=[True, True, True, False],
+    ):
+        super().__init__()
+        print(frag_biases)
+
+        self.pretrained = pretrained
+        self.pretrained2d = pretrained2d
+        self.num_layers = len(depths)
+        self.embed_dim = embed_dim
+        self.patch_norm = patch_norm
+        self.frozen_stages = frozen_stages
+        self.window_size = window_size
+        self.patch_size = patch_size
+
+        # split image into non-overlapping patches
+        self.patch_embed = PatchEmbed3D(
+            patch_size=patch_size,
+            in_chans=in_chans,
+            embed_dim=embed_dim,
+            norm_layer=norm_layer if self.patch_norm else None,
+        )
+
+        self.pos_drop = nn.Dropout(p=drop_rate)
+
+        # stochastic depth
+        dpr = [
+            x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))
+        ]  # stochastic depth decay rule
+
+        # build layers
+        self.layers = nn.ModuleList()
+        for i_layer in range(self.num_layers):
+            layer = BasicLayer(
+                dim=int(embed_dim * 2 ** i_layer),
+                depth=depths[i_layer],
+                num_heads=num_heads[i_layer],
+                window_size=window_size,
+                mlp_ratio=mlp_ratio,
+                qkv_bias=qkv_bias,
+                qk_scale=qk_scale,
+                drop=drop_rate,
+                attn_drop=attn_drop_rate,
+                drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])],
+                norm_layer=norm_layer,
+                downsample=PatchMerging if i_layer < self.num_layers - 1 else None,
+                use_checkpoint=use_checkpoint,
+                jump_attention=jump_attention[i_layer],
+                frag_bias=frag_biases[i_layer],
+            )
+            self.layers.append(layer)
+
+        self.num_features = int(embed_dim * 2 ** (self.num_layers - 1))
+
+        # add a norm layer for each output
+        self.norm = norm_layer(self.num_features)
+
+        self._freeze_stages()
+
+    def _freeze_stages(self):
+        if self.frozen_stages >= 0:
+            self.patch_embed.eval()
+            for param in self.patch_embed.parameters():
+                param.requires_grad = False
+
+        if self.frozen_stages >= 1:
+            self.pos_drop.eval()
+            for i in range(0, self.frozen_stages):
+                m = self.layers[i]
+                m.eval()
+                for param in m.parameters():
+                    param.requires_grad = False
+
+    def inflate_weights(self, logger):
+        """Inflate the swin2d parameters to swin3d.
+
+        The differences between swin3d and swin2d mainly lie in an extra
+        axis. To utilize the pretrained parameters in 2d model,
+        the weight of swin2d models should be inflated to fit in the shapes of
+        the 3d counterpart.
+
+        Args:
+            logger (logging.Logger): The logger used to print
+                debugging infomation.
+        """
+        checkpoint = torch.load(self.pretrained, map_location="cpu")
+        state_dict = checkpoint["model"]
+
+        # delete relative_position_index since we always re-init it
+        relative_position_index_keys = [
+            k for k in state_dict.keys() if "relative_position_index" in k
+        ]
+        for k in relative_position_index_keys:
+            del state_dict[k]
+
+        # delete attn_mask since we always re-init it
+        attn_mask_keys = [k for k in state_dict.keys() if "attn_mask" in k]
+        for k in attn_mask_keys:
+            del state_dict[k]
+
+        state_dict["patch_embed.proj.weight"] = (
+            state_dict["patch_embed.proj.weight"]
+            .unsqueeze(2)
+            .repeat(1, 1, self.patch_size[0], 1, 1)
+            / self.patch_size[0]
+        )
+
+        # bicubic interpolate relative_position_bias_table if not match
+        relative_position_bias_table_keys = [
+            k for k in state_dict.keys() if "relative_position_bias_table" in k
+        ]
+        for k in relative_position_bias_table_keys:
+            relative_position_bias_table_pretrained = state_dict[k]
+            relative_position_bias_table_current = self.state_dict()[k]
+            L1, nH1 = relative_position_bias_table_pretrained.size()
+            L2, nH2 = relative_position_bias_table_current.size()
+            L2 = (2 * self.window_size[1] - 1) * (2 * self.window_size[2] - 1)
+            wd = self.window_size[0]
+            if nH1 != nH2:
+                logger.warning(f"Error in loading {k}, passing")
+            else:
+                if L1 != L2:
+                    S1 = int(L1 ** 0.5)
+                    relative_position_bias_table_pretrained_resized = torch.nn.functional.interpolate(
+                        relative_position_bias_table_pretrained.permute(1, 0).view(
+                            1, nH1, S1, S1
+                        ),
+                        size=(
+                            2 * self.window_size[1] - 1,
+                            2 * self.window_size[2] - 1,
+                        ),
+                        mode="bicubic",
+                    )
+                    relative_position_bias_table_pretrained = relative_position_bias_table_pretrained_resized.view(
+                        nH2, L2
+                    ).permute(
+                        1, 0
+                    )
+            state_dict[k] = relative_position_bias_table_pretrained.repeat(
+                2 * wd - 1, 1
+            )
+
+        msg = self.load_state_dict(state_dict, strict=False)
+        logger.info(msg)
+        logger.info(f"=> loaded successfully '{self.pretrained}'")
+        del checkpoint
+        torch.cuda.empty_cache()
+
+    def load_checkpoint(self, load_path, strict=False):
+        from collections import OrderedDict
+
+        model_state_dict = self.state_dict()
+        state_dict = torch.load(load_path)["state_dict"]
+
+        clean_dict = OrderedDict()
+        for key, value in state_dict.items():
+            if "backbone" in key:
+                clean_key = key[9:]
+                clean_dict[clean_key] = value
+                if "relative_position_bias_table" in clean_key:
+                    forked_key = clean_key.replace(
+                        "relative_position_bias_table", "fragment_position_bias_table"
+                    )
+                    if forked_key in clean_dict:
+                        print(
+                            f"Passing key {forked_key} as it is already in state_dict."
+                        )
+                    else:
+                        clean_dict[forked_key] = value
+
+        ## Only Support for 2X
+        for key, value in model_state_dict.items():
+            if key in clean_dict:
+                if value.shape != clean_dict[key].shape:
+                    clean_dict.pop(key)
+
+        self.load_state_dict(clean_dict, strict=strict)
+
+    def init_weights(self, pretrained=None):
+        print(self.pretrained, self.pretrained2d)
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+
+        def _init_weights(m):
+            if isinstance(m, nn.Linear):
+                trunc_normal_(m.weight, std=0.02)
+                if isinstance(m, nn.Linear) and m.bias is not None:
+                    nn.init.constant_(m.bias, 0)
+            elif isinstance(m, nn.LayerNorm):
+                nn.init.constant_(m.bias, 0)
+                nn.init.constant_(m.weight, 1.0)
+
+        if pretrained:
+            self.pretrained = pretrained
+        if isinstance(self.pretrained, str):
+            self.apply(_init_weights)
+            logger = get_root_logger()
+            logger.info(f"load model from: {self.pretrained}")
+
+            if self.pretrained2d:
+                # Inflate 2D model into 3D model.
+                self.inflate_weights(logger)
+            else:
+                # Directly load 3D model.
+                self.load_checkpoint(self.pretrained, strict=False)  # , logger=logger)
+        elif self.pretrained is None:
+            self.apply(_init_weights)
+        else:
+            raise TypeError("pretrained must be a str or None")
+
+    def forward(self, x, multi=False, require_attn=False):
+        """Forward function."""
+        x = self.patch_embed(x)
+
+        x = self.pos_drop(x)
+
+        if multi:
+            feats = [x]
+
+        for layer in self.layers:
+            x, avg_attns = layer(x.contiguous())
+            if multi:
+                feats += [x]
+
+        x = rearrange(x, "n c d h w -> n d h w c")
+        x = self.norm(x)
+        x = rearrange(x, "n d h w c -> n c d h w")
+
+        if multi:
+            x = feats[:-1] + [x]
+        else:
+            x = x
+
+        if require_attn:
+            return x, avg_attns
+        else:
+            return x
+
+    def train(self, mode=True):
+        """Convert the model into training mode while keep layers freezed."""
+        super(SwinTransformer3D, self).train(mode)
+        self._freeze_stages()

cover/models/backbone_v0_1.py

@@ -0,0 +1,862 @@
+from functools import lru_cache, reduce
+from operator import mul
+
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.utils.checkpoint as checkpoint
+from einops import rearrange
+from timm.models.layers import DropPath, trunc_normal_
+
+
+class Mlp(nn.Module):
+    """Multilayer perceptron."""
+
+    def __init__(
+        self,
+        in_features,
+        hidden_features=None,
+        out_features=None,
+        act_layer=nn.GELU,
+        drop=0.0,
+    ):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.fc1 = nn.Linear(in_features, hidden_features)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features)
+        self.drop = nn.Dropout(drop)
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.drop(x)
+        x = self.fc2(x)
+        x = self.drop(x)
+        return x
+
+
+def window_partition(x, window_size):
+    """
+    Args:
+        x: (B, D, H, W, C)
+        window_size (tuple[int]): window size
+
+    Returns:
+        windows: (B*num_windows, window_size*window_size, C)
+    """
+    B, D, H, W, C = x.shape
+    x = x.view(
+        B,
+        D // window_size[0],
+        window_size[0],
+        H // window_size[1],
+        window_size[1],
+        W // window_size[2],
+        window_size[2],
+        C,
+    )
+    windows = (
+        x.permute(0, 1, 3, 5, 2, 4, 6, 7)
+        .contiguous()
+        .view(-1, reduce(mul, window_size), C)
+    )
+    return windows
+
+
+def window_reverse(windows, window_size, B, D, H, W):
+    """
+    Args:
+        windows: (B*num_windows, window_size, window_size, C)
+        window_size (tuple[int]): Window size
+        H (int): Height of image
+        W (int): Width of image
+
+    Returns:
+        x: (B, D, H, W, C)
+    """
+    x = windows.view(
+        B,
+        D // window_size[0],
+        H // window_size[1],
+        W // window_size[2],
+        window_size[0],
+        window_size[1],
+        window_size[2],
+        -1,
+    )
+    x = x.permute(0, 1, 4, 2, 5, 3, 6, 7).contiguous().view(B, D, H, W, -1)
+    return x
+
+
+def get_window_size(x_size, window_size, shift_size=None):
+    use_window_size = list(window_size)
+    if shift_size is not None:
+        use_shift_size = list(shift_size)
+    for i in range(len(x_size)):
+        if x_size[i] <= window_size[i]:
+            use_window_size[i] = x_size[i]
+            if shift_size is not None:
+                use_shift_size[i] = 0
+
+    if shift_size is None:
+        return tuple(use_window_size)
+    else:
+        return tuple(use_window_size), tuple(use_shift_size)
+
+
+class WindowAttention3D(nn.Module):
+    """Window based multi-head self attention (W-MSA) module with relative position bias.
+    It supports both of shifted and non-shifted window.
+    Args:
+        dim (int): Number of input channels.
+        window_size (tuple[int]): The temporal length, height and width of the window.
+        num_heads (int): Number of attention heads.
+        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set
+        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
+        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
+    """
+
+    def __init__(
+        self,
+        dim,
+        window_size,
+        num_heads,
+        qkv_bias=False,
+        qk_scale=None,
+        attn_drop=0.0,
+        proj_drop=0.0,
+    ):
+
+        super().__init__()
+        self.dim = dim
+        self.window_size = window_size  # Wd, Wh, Ww
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.scale = qk_scale or head_dim ** -0.5
+
+        # define a parameter table of relative position bias
+        self.relative_position_bias_table = nn.Parameter(
+            torch.zeros(
+                (2 * window_size[0] - 1)
+                * (2 * window_size[1] - 1)
+                * (2 * window_size[2] - 1),
+                num_heads,
+            )
+        )  # 2*Wd-1 * 2*Wh-1 * 2*Ww-1, nH
+
+        # get pair-wise relative position index for each token inside the window
+        coords_d = torch.arange(self.window_size[0])
+        coords_h = torch.arange(self.window_size[1])
+        coords_w = torch.arange(self.window_size[2])
+        coords = torch.stack(
+            torch.meshgrid(coords_d, coords_h, coords_w)
+        )  # 3, Wd, Wh, Ww
+        coords_flatten = torch.flatten(coords, 1)  # 3, Wd*Wh*Ww
+        relative_coords = (
+            coords_flatten[:, :, None] - coords_flatten[:, None, :]
+        )  # 3, Wd*Wh*Ww, Wd*Wh*Ww
+        relative_coords = relative_coords.permute(
+            1, 2, 0
+        ).contiguous()  # Wd*Wh*Ww, Wd*Wh*Ww, 3
+        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0
+        relative_coords[:, :, 1] += self.window_size[1] - 1
+        relative_coords[:, :, 2] += self.window_size[2] - 1
+
+        relative_coords[:, :, 0] *= (2 * self.window_size[1] - 1) * (
+            2 * self.window_size[2] - 1
+        )
+        relative_coords[:, :, 1] *= 2 * self.window_size[2] - 1
+        relative_position_index = relative_coords.sum(-1)  # Wd*Wh*Ww, Wd*Wh*Ww
+        self.register_buffer("relative_position_index", relative_position_index)
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(dim, dim)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+        trunc_normal_(self.relative_position_bias_table, std=0.02)
+        self.softmax = nn.Softmax(dim=-1)
+
+    def forward(self, x, mask=None):
+        """Forward function.
+        Args:
+            x: input features with shape of (num_windows*B, N, C)
+            mask: (0/-inf) mask with shape of (num_windows, N, N) or None
+        """
+        B_, N, C = x.shape
+        qkv = (
+            self.qkv(x)
+            .reshape(B_, N, 3, self.num_heads, C // self.num_heads)
+            .permute(2, 0, 3, 1, 4)
+        )
+        q, k, v = qkv[0], qkv[1], qkv[2]  # B_, nH, N, C
+
+        q = q * self.scale
+        attn = q @ k.transpose(-2, -1)
+
+        relative_position_bias = self.relative_position_bias_table[
+            self.relative_position_index[:N, :N].reshape(-1)
+        ].reshape(
+            N, N, -1
+        )  # Wd*Wh*Ww,Wd*Wh*Ww,nH
+        relative_position_bias = relative_position_bias.permute(
+            2, 0, 1
+        ).contiguous()  # nH, Wd*Wh*Ww, Wd*Wh*Ww
+        attn = attn + relative_position_bias.unsqueeze(0)  # B_, nH, N, N
+
+        if mask is not None:
+            nW = mask.shape[0]
+            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(
+                1
+            ).unsqueeze(0)
+            attn = attn.view(-1, self.num_heads, N, N)
+            attn = self.softmax(attn)
+        else:
+            attn = self.softmax(attn)
+
+        attn = self.attn_drop(attn)
+
+        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+        return x
+
+
+class SwinTransformerBlock3D(nn.Module):
+    """Swin Transformer Block.
+
+    Args:
+        dim (int): Number of input channels.
+        num_heads (int): Number of attention heads.
+        window_size (tuple[int]): Window size.
+        shift_size (tuple[int]): Shift size for SW-MSA.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float, optional): Stochastic depth rate. Default: 0.0
+        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+
+    def __init__(
+        self,
+        dim,
+        num_heads,
+        window_size=(2, 7, 7),
+        shift_size=(0, 0, 0),
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        qk_scale=None,
+        drop=0.0,
+        attn_drop=0.0,
+        drop_path=0.0,
+        act_layer=nn.GELU,
+        norm_layer=nn.LayerNorm,
+        use_checkpoint=False,
+        jump_attention=False,
+    ):
+        super().__init__()
+        self.dim = dim
+        self.num_heads = num_heads
+        self.window_size = window_size
+        self.shift_size = shift_size
+        self.mlp_ratio = mlp_ratio
+        self.use_checkpoint = use_checkpoint
+        self.jump_attention = jump_attention
+
+        assert (
+            0 <= self.shift_size[0] < self.window_size[0]
+        ), "shift_size must in 0-window_size"
+        assert (
+            0 <= self.shift_size[1] < self.window_size[1]
+        ), "shift_size must in 0-window_size"
+        assert (
+            0 <= self.shift_size[2] < self.window_size[2]
+        ), "shift_size must in 0-window_size"
+
+        self.norm1 = norm_layer(dim)
+        self.attn = WindowAttention3D(
+            dim,
+            window_size=self.window_size,
+            num_heads=num_heads,
+            qkv_bias=qkv_bias,
+            qk_scale=qk_scale,
+            attn_drop=attn_drop,
+            proj_drop=drop,
+        )
+
+        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
+        self.norm2 = norm_layer(dim)
+        mlp_hidden_dim = int(dim * mlp_ratio)
+        self.mlp = Mlp(
+            in_features=dim,
+            hidden_features=mlp_hidden_dim,
+            act_layer=act_layer,
+            drop=drop,
+        )
+
+    def forward_part1(self, x, mask_matrix):
+        B, D, H, W, C = x.shape
+        window_size, shift_size = get_window_size(
+            (D, H, W), self.window_size, self.shift_size
+        )
+
+        x = self.norm1(x)
+        # pad feature maps to multiples of window size
+        pad_l = pad_t = pad_d0 = 0
+        pad_d1 = (window_size[0] - D % window_size[0]) % window_size[0]
+        pad_b = (window_size[1] - H % window_size[1]) % window_size[1]
+        pad_r = (window_size[2] - W % window_size[2]) % window_size[2]
+        x = F.pad(x, (0, 0, pad_l, pad_r, pad_t, pad_b, pad_d0, pad_d1))
+        _, Dp, Hp, Wp, _ = x.shape
+        # cyclic shift
+        if any(i > 0 for i in shift_size):
+            shifted_x = torch.roll(
+                x,
+                shifts=(-shift_size[0], -shift_size[1], -shift_size[2]),
+                dims=(1, 2, 3),
+            )
+            attn_mask = mask_matrix
+        else:
+            shifted_x = x
+            attn_mask = None
+        # partition windows
+        x_windows = window_partition(shifted_x, window_size)  # B*nW, Wd*Wh*Ww, C
+        # W-MSA/SW-MSA
+        attn_windows = self.attn(x_windows, mask=attn_mask)  # B*nW, Wd*Wh*Ww, C
+        # merge windows
+        attn_windows = attn_windows.view(-1, *(window_size + (C,)))
+        shifted_x = window_reverse(
+            attn_windows, window_size, B, Dp, Hp, Wp
+        )  # B D' H' W' C
+        # reverse cyclic shift
+        if any(i > 0 for i in shift_size):
+            x = torch.roll(
+                shifted_x,
+                shifts=(shift_size[0], shift_size[1], shift_size[2]),
+                dims=(1, 2, 3),
+            )
+        else:
+            x = shifted_x
+
+        if pad_d1 > 0 or pad_r > 0 or pad_b > 0:
+            x = x[:, :D, :H, :W, :].contiguous()
+        return x
+
+    def forward_part2(self, x):
+        return self.drop_path(self.mlp(self.norm2(x)))
+
+    def forward(self, x, mask_matrix):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, D, H, W, C).
+            mask_matrix: Attention mask for cyclic shift.
+        """
+
+        shortcut = x
+        if not self.jump_attention:
+            if self.use_checkpoint:
+                x = checkpoint.checkpoint(self.forward_part1, x, mask_matrix)
+            else:
+                x = self.forward_part1(x, mask_matrix)
+            x = shortcut + self.drop_path(x)
+
+        if self.use_checkpoint:
+            x = x + checkpoint.checkpoint(self.forward_part2, x)
+        else:
+            x = x + self.forward_part2(x)
+
+        return x
+
+
+class PatchMerging(nn.Module):
+    """Patch Merging Layer
+
+    Args:
+        dim (int): Number of input channels.
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+
+    def __init__(self, dim, norm_layer=nn.LayerNorm):
+        super().__init__()
+        self.dim = dim
+        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
+        self.norm = norm_layer(4 * dim)
+
+    def forward(self, x):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, D, H, W, C).
+        """
+        B, D, H, W, C = x.shape
+
+        # padding
+        pad_input = (H % 2 == 1) or (W % 2 == 1)
+        if pad_input:
+            x = F.pad(x, (0, 0, 0, W % 2, 0, H % 2))
+
+        x0 = x[:, :, 0::2, 0::2, :]  # B D H/2 W/2 C
+        x1 = x[:, :, 1::2, 0::2, :]  # B D H/2 W/2 C
+        x2 = x[:, :, 0::2, 1::2, :]  # B D H/2 W/2 C
+        x3 = x[:, :, 1::2, 1::2, :]  # B D H/2 W/2 C
+        x = torch.cat([x0, x1, x2, x3], -1)  # B D H/2 W/2 4*C
+
+        x = self.norm(x)
+        x = self.reduction(x)
+
+        return x
+
+
+# cache each stage results
+@lru_cache()
+def compute_mask(D, H, W, window_size, shift_size, device):
+    img_mask = torch.zeros((1, D, H, W, 1), device=device)  # 1 Dp Hp Wp 1
+    cnt = 0
+    for d in (
+        slice(-window_size[0]),
+        slice(-window_size[0], -shift_size[0]),
+        slice(-shift_size[0], None),
+    ):
+        for h in (
+            slice(-window_size[1]),
+            slice(-window_size[1], -shift_size[1]),
+            slice(-shift_size[1], None),
+        ):
+            for w in (
+                slice(-window_size[2]),
+                slice(-window_size[2], -shift_size[2]),
+                slice(-shift_size[2], None),
+            ):
+                img_mask[:, d, h, w, :] = cnt
+                cnt += 1
+    mask_windows = window_partition(img_mask, window_size)  # nW, ws[0]*ws[1]*ws[2], 1
+    mask_windows = mask_windows.squeeze(-1)  # nW, ws[0]*ws[1]*ws[2]
+    attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
+    attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(
+        attn_mask == 0, float(0.0)
+    )
+    return attn_mask
+
+
+class BasicLayer(nn.Module):
+    """A basic Swin Transformer layer for one stage.
+
+    Args:
+        dim (int): Number of feature channels
+        depth (int): Depths of this stage.
+        num_heads (int): Number of attention head.
+        window_size (tuple[int]): Local window size. Default: (1,7,7).
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
+        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
+        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
+    """
+
+    def __init__(
+        self,
+        dim,
+        depth,
+        num_heads,
+        window_size=(1, 7, 7),
+        mlp_ratio=4.0,
+        qkv_bias=False,
+        qk_scale=None,
+        drop=0.0,
+        attn_drop=0.0,
+        drop_path=0.0,
+        norm_layer=nn.LayerNorm,
+        downsample=None,
+        use_checkpoint=False,
+        jump_attention=False,
+    ):
+        super().__init__()
+        self.window_size = window_size
+        self.shift_size = tuple(i // 2 for i in window_size)
+        self.depth = depth
+        self.use_checkpoint = use_checkpoint
+
+        # build blocks
+        self.blocks = nn.ModuleList(
+            [
+                SwinTransformerBlock3D(
+                    dim=dim,
+                    num_heads=num_heads,
+                    window_size=window_size,
+                    shift_size=(0, 0, 0) if (i % 2 == 0) else self.shift_size,
+                    mlp_ratio=mlp_ratio,
+                    qkv_bias=qkv_bias,
+                    qk_scale=qk_scale,
+                    drop=drop,
+                    attn_drop=attn_drop,
+                    drop_path=drop_path[i]
+                    if isinstance(drop_path, list)
+                    else drop_path,
+                    norm_layer=norm_layer,
+                    use_checkpoint=use_checkpoint,
+                    jump_attention=jump_attention,
+                )
+                for i in range(depth)
+            ]
+        )
+
+        self.downsample = downsample
+        if self.downsample is not None:
+            self.downsample = downsample(dim=dim, norm_layer=norm_layer)
+
+    def forward(self, x):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, C, D, H, W).
+        """
+        # calculate attention mask for SW-MSA
+        B, C, D, H, W = x.shape
+        window_size, shift_size = get_window_size(
+            (D, H, W), self.window_size, self.shift_size
+        )
+        x = rearrange(x, "b c d h w -> b d h w c")
+        Dp = int(np.ceil(D / window_size[0])) * window_size[0]
+        Hp = int(np.ceil(H / window_size[1])) * window_size[1]
+        Wp = int(np.ceil(W / window_size[2])) * window_size[2]
+        attn_mask = compute_mask(Dp, Hp, Wp, window_size, shift_size, x.device)
+        for blk in self.blocks:
+            x = blk(x, attn_mask)
+        x = x.view(B, D, H, W, -1)
+
+        if self.downsample is not None:
+            x = self.downsample(x)
+        x = rearrange(x, "b d h w c -> b c d h w")
+        return x
+
+
+class PatchEmbed3D(nn.Module):
+    """Video to Patch Embedding.
+
+    Args:
+        patch_size (int): Patch token size. Default: (2,4,4).
+        in_chans (int): Number of input video channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        norm_layer (nn.Module, optional): Normalization layer. Default: None
+    """
+
+    def __init__(self, patch_size=(2, 4, 4), in_chans=3, embed_dim=96, norm_layer=None):
+        super().__init__()
+        self.patch_size = patch_size
+
+        self.in_chans = in_chans
+        self.embed_dim = embed_dim
+
+        self.proj = nn.Conv3d(
+            in_chans, embed_dim, kernel_size=patch_size, stride=patch_size
+        )
+        if norm_layer is not None:
+            self.norm = norm_layer(embed_dim)
+        else:
+            self.norm = None
+
+    def forward(self, x):
+        """Forward function."""
+        # padding
+        _, _, D, H, W = x.size()
+        if W % self.patch_size[2] != 0:
+            x = F.pad(x, (0, self.patch_size[2] - W % self.patch_size[2]))
+        if H % self.patch_size[1] != 0:
+            x = F.pad(x, (0, 0, 0, self.patch_size[1] - H % self.patch_size[1]))
+        if D % self.patch_size[0] != 0:
+            x = F.pad(x, (0, 0, 0, 0, 0, self.patch_size[0] - D % self.patch_size[0]))
+
+        x = self.proj(x)  # B C D Wh Ww
+        if self.norm is not None:
+            D, Wh, Ww = x.size(2), x.size(3), x.size(4)
+            x = x.flatten(2).transpose(1, 2)
+            x = self.norm(x)
+            x = x.transpose(1, 2).view(-1, self.embed_dim, D, Wh, Ww)
+
+        return x
+
+
+class SwinTransformer3D(nn.Module):
+    """Swin Transformer backbone.
+        A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows`  -
+          https://arxiv.org/pdf/2103.14030
+
+    Args:
+        patch_size (int | tuple(int)): Patch size. Default: (4,4,4).
+        in_chans (int): Number of input image channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        depths (tuple[int]): Depths of each Swin Transformer stage.
+        num_heads (tuple[int]): Number of attention head of each stage.
+        window_size (int): Window size. Default: 7.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: Truee
+        qk_scale (float): Override default qk scale of head_dim ** -0.5 if set.
+        drop_rate (float): Dropout rate.
+        attn_drop_rate (float): Attention dropout rate. Default: 0.
+        drop_path_rate (float): Stochastic depth rate. Default: 0.2.
+        norm_layer: Normalization layer. Default: nn.LayerNorm.
+        patch_norm (bool): If True, add normalization after patch embedding. Default: False.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters.
+    """
+
+    def __init__(
+        self,
+        pretrained=None,
+        pretrained2d=False,
+        patch_size=(2, 4, 4),
+        in_chans=3,
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=(8, 7, 7),
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        qk_scale=None,
+        drop_rate=0.0,
+        attn_drop_rate=0.0,
+        drop_path_rate=0.1,
+        norm_layer=nn.LayerNorm,
+        patch_norm=True,
+        frozen_stages=-1,
+        use_checkpoint=True,
+        jump_attention=[False, False, False, False],
+    ):
+        super().__init__()
+
+        self.pretrained = pretrained
+        self.pretrained2d = pretrained2d
+        self.num_layers = len(depths)
+        self.embed_dim = embed_dim
+        self.patch_norm = patch_norm
+        self.frozen_stages = frozen_stages
+        self.window_size = window_size
+        self.patch_size = patch_size
+
+        # split image into non-overlapping patches
+        self.patch_embed = PatchEmbed3D(
+            patch_size=patch_size,
+            in_chans=in_chans,
+            embed_dim=embed_dim,
+            norm_layer=norm_layer if self.patch_norm else None,
+        )
+
+        self.pos_drop = nn.Dropout(p=drop_rate)
+
+        # stochastic depth
+        dpr = [
+            x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))
+        ]  # stochastic depth decay rule
+
+        # build layers
+        self.layers = nn.ModuleList()
+        for i_layer in range(self.num_layers):
+            layer = BasicLayer(
+                dim=int(embed_dim * 2 ** i_layer),
+                depth=depths[i_layer],
+                num_heads=num_heads[i_layer],
+                window_size=window_size,
+                mlp_ratio=mlp_ratio,
+                qkv_bias=qkv_bias,
+                qk_scale=qk_scale,
+                drop=drop_rate,
+                attn_drop=attn_drop_rate,
+                drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])],
+                norm_layer=norm_layer,
+                downsample=PatchMerging if i_layer < self.num_layers - 1 else None,
+                use_checkpoint=use_checkpoint,
+                jump_attention=jump_attention[i_layer],
+            )
+            self.layers.append(layer)
+
+        self.num_features = int(embed_dim * 2 ** (self.num_layers - 1))
+
+        # add a norm layer for each output
+        self.norm = norm_layer(self.num_features)
+
+        self._freeze_stages()
+
+    def _freeze_stages(self):
+        if self.frozen_stages >= 0:
+            self.patch_embed.eval()
+            for param in self.patch_embed.parameters():
+                param.requires_grad = False
+
+        if self.frozen_stages >= 1:
+            self.pos_drop.eval()
+            for i in range(0, self.frozen_stages):
+                m = self.layers[i]
+                m.eval()
+                for param in m.parameters():
+                    param.requires_grad = False
+
+    def inflate_weights(self, logger):
+        """Inflate the swin2d parameters to swin3d.
+
+        The differences between swin3d and swin2d mainly lie in an extra
+        axis. To utilize the pretrained parameters in 2d model,
+        the weight of swin2d models should be inflated to fit in the shapes of
+        the 3d counterpart.
+
+        Args:
+            logger (logging.Logger): The logger used to print
+                debugging infomation.
+        """
+        checkpoint = torch.load(self.pretrained, map_location="cpu")
+        state_dict = checkpoint["model"]
+
+        # delete relative_position_index since we always re-init it
+        relative_position_index_keys = [
+            k for k in state_dict.keys() if "relative_position_index" in k
+        ]
+        for k in relative_position_index_keys:
+            del state_dict[k]
+
+        # delete attn_mask since we always re-init it
+        attn_mask_keys = [k for k in state_dict.keys() if "attn_mask" in k]
+        for k in attn_mask_keys:
+            del state_dict[k]
+
+        state_dict["patch_embed.proj.weight"] = (
+            state_dict["patch_embed.proj.weight"]
+            .unsqueeze(2)
+            .repeat(1, 1, self.patch_size[0], 1, 1)
+            / self.patch_size[0]
+        )
+
+        # bicubic interpolate relative_position_bias_table if not match
+        relative_position_bias_table_keys = [
+            k for k in state_dict.keys() if "relative_position_bias_table" in k
+        ]
+        for k in relative_position_bias_table_keys:
+            relative_position_bias_table_pretrained = state_dict[k]
+            relative_position_bias_table_current = self.state_dict()[k]
+            L1, nH1 = relative_position_bias_table_pretrained.size()
+            L2, nH2 = relative_position_bias_table_current.size()
+            L2 = (2 * self.window_size[1] - 1) * (2 * self.window_size[2] - 1)
+            wd = self.window_size[0]
+            if nH1 != nH2:
+                logger.warning(f"Error in loading {k}, passing")
+            else:
+                if L1 != L2:
+                    S1 = int(L1 ** 0.5)
+                    relative_position_bias_table_pretrained_resized = torch.nn.functional.interpolate(
+                        relative_position_bias_table_pretrained.permute(1, 0).view(
+                            1, nH1, S1, S1
+                        ),
+                        size=(
+                            2 * self.window_size[1] - 1,
+                            2 * self.window_size[2] - 1,
+                        ),
+                        mode="bicubic",
+                    )
+                    relative_position_bias_table_pretrained = relative_position_bias_table_pretrained_resized.view(
+                        nH2, L2
+                    ).permute(
+                        1, 0
+                    )
+            state_dict[k] = relative_position_bias_table_pretrained.repeat(
+                2 * wd - 1, 1
+            )
+
+        msg = self.load_state_dict(state_dict, strict=False)
+        logger.info(msg)
+        logger.info(f"=> loaded successfully '{self.pretrained}'")
+        del checkpoint
+        torch.cuda.empty_cache()
+
+    def load_checkpoint(self, load_path, strict=False):
+        from collections import OrderedDict
+
+        model_state_dict = self.state_dict()
+        state_dict = torch.load(load_path)
+        if "state_dict" in state_dict.keys():
+            state_dict = state_dict["state_dict"]
+
+        clean_dict = OrderedDict()
+        for key, value in state_dict.items():
+            if "backbone" in key:
+                clean_key = key[9:]
+                clean_dict[clean_key] = value
+
+        if not strict:
+            for key, value in model_state_dict.items():
+                if key in clean_dict:
+                    if value.shape != clean_dict[key].shape:
+                        clean_dict.pop(key)
+
+        self.load_state_dict(clean_dict, strict=strict)
+
+    def init_weights(self, pretrained=None):
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+
+        def _init_weights(m):
+            if isinstance(m, nn.Linear):
+                trunc_normal_(m.weight, std=0.02)
+                if isinstance(m, nn.Linear) and m.bias is not None:
+                    nn.init.constant_(m.bias, 0)
+            elif isinstance(m, nn.LayerNorm):
+                nn.init.constant_(m.bias, 0)
+                nn.init.constant_(m.weight, 1.0)
+
+        if pretrained:
+            self.pretrained = pretrained
+        if isinstance(self.pretrained, str):
+            self.apply(_init_weights)
+            logger = get_root_logger()
+            logger.info(f"load model from: {self.pretrained}")
+
+            if self.pretrained2d:
+                # Inflate 2D model into 3D model.
+                self.inflate_weights(logger)
+            else:
+                # Directly load 3D model.
+                self.load_checkpoint(self.pretrained, strict=False)  # , logger=logger)
+        elif self.pretrained is None:
+            self.apply(_init_weights)
+        else:
+            raise TypeError("pretrained must be a str or None")
+
+    def forward(self, x, multi=False):
+        """Forward function."""
+        x = self.patch_embed(x)
+
+        x = self.pos_drop(x)
+
+        if multi:
+            feats = [x]
+
+        for layer in self.layers:
+            x = layer(x.contiguous())
+            if multi:
+                feats += [x]
+
+        x = rearrange(x, "n c d h w -> n d h w c")
+        x = self.norm(x)
+        x = rearrange(x, "n d h w c -> n c d h w")
+
+        if multi:
+            return feats[:-1] + [x]
+        else:
+            return x
+
+    def train(self, mode=True):
+        """Convert the model into training mode while keep layers freezed."""
+        super(SwinTransformer3D, self).train(mode)
+        self._freeze_stages()

cover/models/clip_model.py

@@ -0,0 +1,640 @@
+import hashlib
+import os
+import urllib
+import warnings
+from tqdm import tqdm
+from typing import Tuple, Union, List
+from collections import OrderedDict
+
+import numpy as np
+import torch
+import torch.nn.functional as F
+from torch import nn
+
+
+_MODELS = {
+    "RN50": "https://openaipublic.azureedge.net/clip/models/afeb0e10f9e5a86da6080e35cf09123aca3b358a0c3e3b6c78a7b63bc04b6762/RN50.pt",
+    "RN101": "https://openaipublic.azureedge.net/clip/models/8fa8567bab74a42d41c5915025a8e4538c3bdbe8804a470a72f30b0d94fab599/RN101.pt",
+    "RN50x4": "https://openaipublic.azureedge.net/clip/models/7e526bd135e493cef0776de27d5f42653e6b4c8bf9e0f653bb11773263205fdd/RN50x4.pt",
+    "RN50x16": "https://openaipublic.azureedge.net/clip/models/52378b407f34354e150460fe41077663dd5b39c54cd0bfd2b27167a4a06ec9aa/RN50x16.pt",
+    "RN50x64": "https://openaipublic.azureedge.net/clip/models/be1cfb55d75a9666199fb2206c106743da0f6468c9d327f3e0d0a543a9919d9c/RN50x64.pt",
+    "ViT-B/32": "https://openaipublic.azureedge.net/clip/models/40d365715913c9da98579312b702a82c18be219cc2a73407c4526f58eba950af/ViT-B-32.pt",
+    "ViT-B/16": "https://openaipublic.azureedge.net/clip/models/5806e77cd80f8b59890b7e101eabd078d9fb84e6937f9e85e4ecb61988df416f/ViT-B-16.pt",
+    "ViT-L/14": "https://openaipublic.azureedge.net/clip/models/b8cca3fd41ae0c99ba7e8951adf17d267cdb84cd88be6f7c2e0eca1737a03836/ViT-L-14.pt",
+    "ViT-L/14@336px": "https://openaipublic.azureedge.net/clip/models/3035c92b350959924f9f00213499208652fc7ea050643e8b385c2dac08641f02/ViT-L-14-336px.pt",
+}
+
+
+def _download(url: str, root: str):
+    os.makedirs(root, exist_ok=True)
+    filename = os.path.basename(url)
+
+    expected_sha256 = url.split("/")[-2]
+    download_target = os.path.join(root, filename)
+
+    if os.path.exists(download_target) and not os.path.isfile(download_target):
+        raise RuntimeError(f"{download_target} exists and is not a regular file")
+
+    if os.path.isfile(download_target):
+        if hashlib.sha256(open(download_target, "rb").read()).hexdigest() == expected_sha256:
+            return download_target
+        else:
+            warnings.warn(f"{download_target} exists, but the SHA256 checksum does not match; re-downloading the file")
+
+    with urllib.request.urlopen(url) as source, open(download_target, "wb") as output:
+        with tqdm(total=int(source.info().get("Content-Length")), ncols=80, unit='iB', unit_scale=True, unit_divisor=1024) as loop:
+            while True:
+                buffer = source.read(8192)
+                if not buffer:
+                    break
+
+                output.write(buffer)
+                loop.update(len(buffer))
+
+    if hashlib.sha256(open(download_target, "rb").read()).hexdigest() != expected_sha256:
+        raise RuntimeError("Model has been downloaded but the SHA256 checksum does not not match")
+
+    return download_target
+
+
+def available_models() -> List[str]:
+    """Returns the names of available CLIP models"""
+    return list(_MODELS.keys())
+
+
+def load(name: str, device: Union[str, torch.device] = "cuda" if torch.cuda.is_available() else "cpu", jit: bool = False, download_root: str = None):
+    """Load a CLIP model
+    Parameters
+    ----------
+    name : str
+        A model name listed by `clip.available_models()`, or the path to a model checkpoint containing the state_dict
+    device : Union[str, torch.device]
+        The device to put the loaded model
+    jit : bool
+        Whether to load the optimized JIT model or more hackable non-JIT model (default).
+    download_root: str
+        path to download the model files; by default, it uses "~/.cache/clip"
+    Returns
+    -------
+    model : torch.nn.Module
+        The CLIP model
+    preprocess : Callable[[PIL.Image], torch.Tensor]
+        A torchvision transform that converts a PIL image into a tensor that the returned model can take as its input
+    """
+    if name in _MODELS:
+        model_path = _download(_MODELS[name], download_root or os.path.expanduser("~/.cache/clip"))
+    elif os.path.isfile(name):
+        model_path = name
+    else:
+        raise RuntimeError(f"Model {name} not found; available models = {available_models()}")
+
+    with open(model_path, 'rb') as opened_file:
+        try:
+            # loading JIT archive
+            model = torch.jit.load(opened_file, map_location=device if jit else "cpu").eval()
+            state_dict = None
+        except RuntimeError:
+            # loading saved state dict
+            if jit:
+                warnings.warn(f"File {model_path} is not a JIT archive. Loading as a state dict instead")
+                jit = False
+            state_dict = torch.load(opened_file, map_location="cpu")
+
+    if not jit:
+        model = build_model(state_dict or model.state_dict()).to(device)
+        if str(device) == "cpu":
+            model.float()
+        return model
+
+    # patch the device names
+    device_holder = torch.jit.trace(lambda: torch.ones([]).to(torch.device(device)), example_inputs=[])
+    device_node = [n for n in device_holder.graph.findAllNodes("prim::Constant") if "Device" in repr(n)][-1]
+
+    def patch_device(module):
+        try:
+            graphs = [module.graph] if hasattr(module, "graph") else []
+        except RuntimeError:
+            graphs = []
+
+        if hasattr(module, "forward1"):
+            graphs.append(module.forward1.graph)
+
+        for graph in graphs:
+            for node in graph.findAllNodes("prim::Constant"):
+                if "value" in node.attributeNames() and str(node["value"]).startswith("cuda"):
+                    node.copyAttributes(device_node)
+
+    model.apply(patch_device)
+    patch_device(model.encode_image)
+    patch_device(model.encode_text)
+
+    # patch dtype to float32 on CPU
+    if str(device) == "cpu":
+        float_holder = torch.jit.trace(lambda: torch.ones([]).float(), example_inputs=[])
+        float_input = list(float_holder.graph.findNode("aten::to").inputs())[1]
+        float_node = float_input.node()
+
+        def patch_float(module):
+            try:
+                graphs = [module.graph] if hasattr(module, "graph") else []
+            except RuntimeError:
+                graphs = []
+
+            if hasattr(module, "forward1"):
+                graphs.append(module.forward1.graph)
+
+            for graph in graphs:
+                for node in graph.findAllNodes("aten::to"):
+                    inputs = list(node.inputs())
+                    for i in [1, 2]:  # dtype can be the second or third argument to aten::to()
+                        if inputs[i].node()["value"] == 5:
+                            inputs[i].node().copyAttributes(float_node)
+
+        model.apply(patch_float)
+        patch_float(model.encode_image)
+        patch_float(model.encode_text)
+
+        model.float()
+
+    return model
+
+
+class Bottleneck(nn.Module):
+    expansion = 4
+
+    def __init__(self, inplanes, planes, stride=1):
+        super().__init__()
+
+        # all conv layers have stride 1. an avgpool is performed after the second convolution when stride > 1
+        self.conv1 = nn.Conv2d(inplanes, planes, 1, bias=False)
+        self.bn1 = nn.BatchNorm2d(planes)
+
+        self.conv2 = nn.Conv2d(planes, planes, 3, padding=1, bias=False)
+        self.bn2 = nn.BatchNorm2d(planes)
+
+        self.avgpool = nn.AvgPool2d(stride) if stride > 1 else nn.Identity()
+
+        self.conv3 = nn.Conv2d(planes, planes * self.expansion, 1, bias=False)
+        self.bn3 = nn.BatchNorm2d(planes * self.expansion)
+
+        self.relu = nn.ReLU(inplace=True)
+        self.downsample = None
+        self.stride = stride
+
+        if stride > 1 or inplanes != planes * Bottleneck.expansion:
+            # downsampling layer is prepended with an avgpool, and the subsequent convolution has stride 1
+            self.downsample = nn.Sequential(OrderedDict([
+                ("-1", nn.AvgPool2d(stride)),
+                ("0", nn.Conv2d(inplanes, planes * self.expansion, 1, stride=1, bias=False)),
+                ("1", nn.BatchNorm2d(planes * self.expansion))
+            ]))
+
+    def forward(self, x: torch.Tensor):
+        identity = x
+
+        out = self.relu(self.bn1(self.conv1(x)))
+        out = self.relu(self.bn2(self.conv2(out)))
+        out = self.avgpool(out)
+        out = self.bn3(self.conv3(out))
+
+        if self.downsample is not None:
+            identity = self.downsample(x)
+
+        out += identity
+        out = self.relu(out)
+        return out
+
+
+class AttentionPool2d(nn.Module):
+    def __init__(self, spacial_dim: int, embed_dim: int, num_heads: int, output_dim: int = None):
+        super().__init__()
+        self.positional_embedding = nn.Parameter(torch.randn(spacial_dim ** 2 + 1, embed_dim) / embed_dim ** 0.5)
+        self.k_proj = nn.Linear(embed_dim, embed_dim)
+        self.q_proj = nn.Linear(embed_dim, embed_dim)
+        self.v_proj = nn.Linear(embed_dim, embed_dim)
+        self.c_proj = nn.Linear(embed_dim, output_dim or embed_dim)
+        self.num_heads = num_heads
+        self.spacial_dim = spacial_dim
+        self.embed_dim = embed_dim
+
+    def forward(self, x, return_token=False, pos_embedding=False):
+        n, c, h, w = x.shape
+        x = x.reshape(x.shape[0], x.shape[1], x.shape[2] * x.shape[3]).permute(2, 0, 1)  # NCHW -> (HW)NC
+        x = torch.cat([x.mean(dim=0, keepdim=True), x], dim=0)  # (HW+1)NC
+        if pos_embedding:
+            positional_embedding_resize = F.interpolate(self.positional_embedding.unsqueeze(
+                0).unsqueeze(0), size=(x.size(0), x.size(2)), mode='bicubic').squeeze(0).squeeze(0)
+            x = x + positional_embedding_resize[:, None, :].to(x.dtype)  # (HW+1)NC
+
+        x, _ = F.multi_head_attention_forward(
+            query=x, key=x, value=x,
+            embed_dim_to_check=x.shape[-1],
+            num_heads=self.num_heads,
+            q_proj_weight=self.q_proj.weight,
+            k_proj_weight=self.k_proj.weight,
+            v_proj_weight=self.v_proj.weight,
+            in_proj_weight=None,
+            in_proj_bias=torch.cat([self.q_proj.bias, self.k_proj.bias, self.v_proj.bias]),
+            bias_k=None,
+            bias_v=None,
+            add_zero_attn=False,
+            dropout_p=0,
+            out_proj_weight=self.c_proj.weight,
+            out_proj_bias=self.c_proj.bias,
+            use_separate_proj_weight=True,
+            training=self.training,
+            need_weights=False
+        )
+
+        if return_token:
+            return x[0], x[1:]
+        else:
+            return x[0]
+
+
+class ModifiedResNet(nn.Module):
+    """
+    A ResNet class that is similar to torchvision's but contains the following changes:
+    - There are now 3 "stem" convolutions as opposed to 1, with an average pool instead of a max pool.
+    - Performs anti-aliasing strided convolutions, where an avgpool is prepended to convolutions with stride > 1
+    - The final pooling layer is a QKV attention instead of an average pool
+    """
+
+    def __init__(self, layers, output_dim, heads, input_resolution=224, width=64):
+        super().__init__()
+        self.output_dim = output_dim
+        self.input_resolution = input_resolution
+
+        # the 3-layer stem
+        self.conv1 = nn.Conv2d(3, width // 2, kernel_size=3, stride=2, padding=1, bias=False)
+        self.bn1 = nn.BatchNorm2d(width // 2)
+        self.conv2 = nn.Conv2d(width // 2, width // 2, kernel_size=3, padding=1, bias=False)
+        self.bn2 = nn.BatchNorm2d(width // 2)
+        self.conv3 = nn.Conv2d(width // 2, width, kernel_size=3, padding=1, bias=False)
+        self.bn3 = nn.BatchNorm2d(width)
+        self.avgpool = nn.AvgPool2d(2)
+        self.relu = nn.ReLU(inplace=True)
+
+        # residual layers
+        self._inplanes = width  # this is a *mutable* variable used during construction
+        self.layer1 = self._make_layer(width, layers[0])
+        self.layer2 = self._make_layer(width * 2, layers[1], stride=2)
+        self.layer3 = self._make_layer(width * 4, layers[2], stride=2)
+        self.layer4 = self._make_layer(width * 8, layers[3], stride=2)
+
+        self.feature_dim_list = [width, width * 4, width * 8, width * 16, width * 32]
+
+        embed_dim = width * 32  # the ResNet feature dimension
+        self.attnpool = AttentionPool2d(input_resolution // 32, embed_dim, heads, output_dim)
+
+    def _make_layer(self, planes, blocks, stride=1):
+        layers = [Bottleneck(self._inplanes, planes, stride)]
+
+        self._inplanes = planes * Bottleneck.expansion
+        for _ in range(1, blocks):
+            layers.append(Bottleneck(self._inplanes, planes))
+
+        return nn.Sequential(*layers)
+    
+    def forward_features(self, x, return_token=False, pos_embedding=False):
+        def stem(x):
+            for conv, bn in [(self.conv1, self.bn1), (self.conv2, self.bn2), (self.conv3, self.bn3)]:
+                x = self.relu(bn(conv(x)))
+            x = self.avgpool(x)
+            return x
+        
+        x = x.type(self.conv1.weight.dtype)
+        x = stem(x)
+        feat_list = [x]
+        x = self.layer1(x)
+        feat_list += [x]
+        x = self.layer2(x)
+        feat_list += [x]
+        x = self.layer3(x)
+        feat_list += [x]
+        x = self.layer4(x)
+        feat_list += [x]
+        return feat_list
+
+    def forward(self, x, return_token=False, pos_embedding=False):
+        def stem(x):
+            for conv, bn in [(self.conv1, self.bn1), (self.conv2, self.bn2), (self.conv3, self.bn3)]:
+                x = self.relu(bn(conv(x)))
+            x = self.avgpool(x)
+            return x
+
+        x = x.type(self.conv1.weight.dtype)
+        x = stem(x)
+        x = self.layer1(x)
+        x = self.layer2(x)
+        x = self.layer3(x)
+        x = self.layer4(x)
+
+        if return_token:
+            x, tokens = self.attnpool(x, return_token, pos_embedding)
+            return x, tokens
+        else:
+            x = self.attnpool(x, return_token, pos_embedding)
+            return x
+
+
+class LayerNorm(nn.LayerNorm):
+    """Subclass torch's LayerNorm to handle fp16."""
+
+    def forward(self, x: torch.Tensor):
+        orig_type = x.dtype
+        ret = super().forward(x.type(torch.float32))
+        return ret.type(orig_type)
+
+
+class QuickGELU(nn.Module):
+    def forward(self, x: torch.Tensor):
+        return x * torch.sigmoid(1.702 * x)
+
+
+class ResidualAttentionBlock(nn.Module):
+    def __init__(self, d_model: int, n_head: int, attn_mask: torch.Tensor = None):
+        super().__init__()
+
+        self.attn = nn.MultiheadAttention(d_model, n_head)
+        self.ln_1 = LayerNorm(d_model)
+        self.mlp = nn.Sequential(OrderedDict([
+            ("c_fc", nn.Linear(d_model, d_model * 4)),
+            ("gelu", QuickGELU()),
+            ("c_proj", nn.Linear(d_model * 4, d_model))
+        ]))
+        self.ln_2 = LayerNorm(d_model)
+        self.attn_mask = attn_mask
+
+    def attention(self, x: torch.Tensor):
+        self.attn_mask = self.attn_mask.to(dtype=x.dtype, device=x.device) if self.attn_mask is not None else None
+        return self.attn(x, x, x, need_weights=False, attn_mask=self.attn_mask)[0]
+
+    def forward(self, x: torch.Tensor):
+        x = x + self.attention(self.ln_1(x))
+        x = x + self.mlp(self.ln_2(x))
+        return x
+
+
+class Transformer(nn.Module):
+    def __init__(self, width: int, layers: int, heads: int, attn_mask: torch.Tensor = None):
+        super().__init__()
+        self.width = width
+        self.layers = layers
+        self.resblocks = nn.Sequential(*[ResidualAttentionBlock(width, heads, attn_mask) for _ in range(layers)])
+
+    def forward(self, x: torch.Tensor):
+        return self.resblocks(x)
+
+
+class VisionTransformer(nn.Module):
+    def __init__(self, input_resolution: int, patch_size: int, width: int, layers: int, heads: int, output_dim: int):
+        super().__init__()
+        self.input_resolution = input_resolution
+        self.output_dim = output_dim
+        self.conv1 = nn.Conv2d(in_channels=3, out_channels=width,
+                               kernel_size=patch_size, stride=patch_size, bias=False)
+
+        scale = width ** -0.5
+        self.class_embedding = nn.Parameter(scale * torch.randn(width))
+        self.positional_embedding = nn.Parameter(scale * torch.randn((input_resolution // patch_size) ** 2 + 1, width))
+        self.ln_pre = LayerNorm(width)
+
+        self.transformer = Transformer(width, layers, heads)
+
+        self.ln_post = LayerNorm(width)
+        self.proj = nn.Parameter(scale * torch.randn(width, output_dim))
+
+    def forward(self, x: torch.Tensor, return_token=True, pos_embedding=False):
+        x = self.conv1(x)  # shape = [*, width, grid, grid]
+        x = x.reshape(x.shape[0], x.shape[1], -1)  # shape = [*, width, grid ** 2]
+        x = x.permute(0, 2, 1)  # shape = [*, grid ** 2, width]
+        x = torch.cat([self.class_embedding.to(x.dtype) + torch.zeros(x.shape[0], 1, x.shape[-1],
+                                                                      dtype=x.dtype, device=x.device), x], dim=1)  # shape = [*, grid ** 2 + 1, width]
+
+        if pos_embedding:
+            positional_embedding_resize = F.interpolate(self.positional_embedding.unsqueeze(
+                0).unsqueeze(0), size=(x.size(1), x.size(2)), mode='bicubic').squeeze(0).squeeze(0)
+            x = x + positional_embedding_resize.to(x.dtype)
+
+        x = self.ln_pre(x)
+
+        x = x.permute(1, 0, 2)  # NLD -> LND
+        x = self.transformer(x)
+        x = x.permute(1, 0, 2)  # LND -> NLD
+
+        token = self.ln_post(x[:, 1:, :])
+
+        x = self.ln_post(x[:, 0, :])
+
+        if self.proj is not None:
+            x = x @ self.proj
+
+        if return_token:
+            return x, token
+        else:
+            return x
+
+
+class CLIP(nn.Module):
+    def __init__(self,
+                 embed_dim: int,
+                 # vision
+                 image_resolution: int,
+                 vision_layers: Union[Tuple[int, int, int, int], int],
+                 vision_width: int,
+                 vision_patch_size: int,
+                 # text
+                 context_length: int,
+                 vocab_size: int,
+                 transformer_width: int,
+                 transformer_heads: int,
+                 transformer_layers: int
+                 ):
+        super().__init__()
+
+        self.context_length = context_length
+
+        if isinstance(vision_layers, (tuple, list)):
+            vision_heads = vision_width * 32 // 64
+            self.visual = ModifiedResNet(
+                layers=vision_layers,
+                output_dim=embed_dim,
+                heads=vision_heads,
+                input_resolution=image_resolution,
+                width=vision_width
+            )
+        else:
+            vision_heads = vision_width // 64
+            self.visual = VisionTransformer(
+                input_resolution=image_resolution,
+                patch_size=vision_patch_size,
+                width=vision_width,
+                layers=vision_layers,
+                heads=vision_heads,
+                output_dim=embed_dim
+            )
+
+        self.transformer = Transformer(
+            width=transformer_width,
+            layers=transformer_layers,
+            heads=transformer_heads,
+            attn_mask=self.build_attention_mask()
+        )
+
+        self.vocab_size = vocab_size
+        self.token_embedding = nn.Embedding(vocab_size, transformer_width)
+        self.positional_embedding = nn.Parameter(torch.empty(self.context_length, transformer_width))
+        self.ln_final = LayerNorm(transformer_width)
+
+        self.text_projection = nn.Parameter(torch.empty(transformer_width, embed_dim))
+        self.logit_scale = nn.Parameter(torch.ones([]) * np.log(1 / 0.07))
+
+        self.initialize_parameters()
+
+    def initialize_parameters(self):
+        nn.init.normal_(self.token_embedding.weight, std=0.02)
+        nn.init.normal_(self.positional_embedding, std=0.01)
+
+        if isinstance(self.visual, ModifiedResNet):
+            if self.visual.attnpool is not None:
+                std = self.visual.attnpool.c_proj.in_features ** -0.5
+                nn.init.normal_(self.visual.attnpool.q_proj.weight, std=std)
+                nn.init.normal_(self.visual.attnpool.k_proj.weight, std=std)
+                nn.init.normal_(self.visual.attnpool.v_proj.weight, std=std)
+                nn.init.normal_(self.visual.attnpool.c_proj.weight, std=std)
+
+            for resnet_block in [self.visual.layer1, self.visual.layer2, self.visual.layer3, self.visual.layer4]:
+                for name, param in resnet_block.named_parameters():
+                    if name.endswith("bn3.weight"):
+                        nn.init.zeros_(param)
+
+        proj_std = (self.transformer.width ** -0.5) * ((2 * self.transformer.layers) ** -0.5)
+        attn_std = self.transformer.width ** -0.5
+        fc_std = (2 * self.transformer.width) ** -0.5
+        for block in self.transformer.resblocks:
+            nn.init.normal_(block.attn.in_proj_weight, std=attn_std)
+            nn.init.normal_(block.attn.out_proj.weight, std=proj_std)
+            nn.init.normal_(block.mlp.c_fc.weight, std=fc_std)
+            nn.init.normal_(block.mlp.c_proj.weight, std=proj_std)
+
+        if self.text_projection is not None:
+            nn.init.normal_(self.text_projection, std=self.transformer.width ** -0.5)
+
+    def build_attention_mask(self):
+        # lazily create causal attention mask, with full attention between the vision tokens
+        # pytorch uses additive attention mask; fill with -inf
+        mask = torch.empty(self.context_length, self.context_length)
+        mask.fill_(float("-inf"))
+        mask.triu_(1)  # zero out the lower diagonal
+        return mask
+
+    @property
+    def dtype(self):
+        return self.visual.conv1.weight.dtype
+
+    def encode_image(self, image, pos_embedding):
+        return self.visual(image.type(self.dtype), pos_embedding=pos_embedding)
+
+    def encode_text(self, text):
+        x = self.token_embedding(text).type(self.dtype)  # [batch_size, n_ctx, d_model]
+
+        x = x + self.positional_embedding.type(self.dtype)
+        x = x.permute(1, 0, 2)  # NLD -> LND
+        x = self.transformer(x)
+        x = x.permute(1, 0, 2)  # LND -> NLD
+        x = self.ln_final(x).type(self.dtype)
+
+        # x.shape = [batch_size, n_ctx, transformer.width]
+        # take features from the eot embedding (eot_token is the highest number in each sequence)
+        x = x[torch.arange(x.shape[0]), text.argmax(dim=-1)] @ self.text_projection
+
+        return x
+
+    def forward(self, image, text, pos_embedding=False, text_features=None):
+        # only use the image encoder in CLIP
+        image_features, token_features = self.encode_image(image, pos_embedding)
+        
+        # normalized features
+        image_features = image_features / image_features.norm(dim=-1, keepdim=True)
+
+        # don't process encode_text
+        # if text_features is None:
+        #     text_features = self.encode_text(text)
+        # text_features = text_features / text_features.norm(dim=-1, keepdim=True)
+
+        # cosine similarity as logits
+        # logit_scale = self.logit_scale.exp()
+        # logits_per_image = logit_scale * image_features @ text_features.t()
+        # logits_per_text = logits_per_image.t()
+        logits_per_image = 0
+        logits_per_text = 0
+
+        # shape = [global_batch_size, global_batch_size]
+        return logits_per_image, logits_per_text, image_features, token_features
+
+
+def convert_weights(model: nn.Module):
+    """Convert applicable model parameters to fp16"""
+
+    def _convert_weights_to_fp16(l):
+        if isinstance(l, (nn.Conv1d, nn.Conv2d, nn.Linear)):
+            l.weight.data = l.weight.data.half()
+            if l.bias is not None:
+                l.bias.data = l.bias.data.half()
+
+        if isinstance(l, nn.MultiheadAttention):
+            for attr in [*[f"{s}_proj_weight" for s in ["in", "q", "k", "v"]], "in_proj_bias", "bias_k", "bias_v"]:
+                tensor = getattr(l, attr)
+                if tensor is not None:
+                    tensor.data = tensor.data.half()
+
+        for name in ["text_projection", "proj"]:
+            if hasattr(l, name):
+                attr = getattr(l, name)
+                if attr is not None:
+                    attr.data = attr.data.half()
+
+    model.apply(_convert_weights_to_fp16)
+
+
+def build_model(state_dict: dict):
+    vit = "visual.proj" in state_dict
+
+    if vit:
+        vision_width = state_dict["visual.conv1.weight"].shape[0]
+        vision_layers = len([k for k in state_dict.keys() if k.startswith(
+            "visual.") and k.endswith(".attn.in_proj_weight")])
+        vision_patch_size = state_dict["visual.conv1.weight"].shape[-1]
+        grid_size = round((state_dict["visual.positional_embedding"].shape[0] - 1) ** 0.5)
+        image_resolution = vision_patch_size * grid_size
+    else:
+        counts: list = [len(set(k.split(".")[2]
+                            for k in state_dict if k.startswith(f"visual.layer{b}"))) for b in [1, 2, 3, 4]]
+        vision_layers = tuple(counts)
+        vision_width = state_dict["visual.layer1.0.conv1.weight"].shape[0]
+        output_width = round((state_dict["visual.attnpool.positional_embedding"].shape[0] - 1) ** 0.5)
+        vision_patch_size = None
+        assert output_width ** 2 + 1 == state_dict["visual.attnpool.positional_embedding"].shape[0]
+        image_resolution = output_width * 32
+
+    embed_dim = state_dict["text_projection"].shape[1]
+    context_length = state_dict["positional_embedding"].shape[0]
+    vocab_size = state_dict["token_embedding.weight"].shape[0]
+    transformer_width = state_dict["ln_final.weight"].shape[0]
+    transformer_heads = transformer_width // 64
+    transformer_layers = len(set(k.split(".")[2] for k in state_dict if k.startswith(f"transformer.resblocks")))
+
+    model = CLIP(
+        embed_dim,
+        image_resolution, vision_layers, vision_width, vision_patch_size,
+        context_length, vocab_size, transformer_width, transformer_heads, transformer_layers
+    )
+
+    for key in ["input_resolution", "context_length", "vocab_size"]:
+        if key in state_dict:
+            del state_dict[key]
+
+    convert_weights(model)
+    model.load_state_dict(state_dict)
+    return model.eval()

cover/models/clipiqa_arch.py

@@ -0,0 +1,165 @@
+r"""CLIP-IQA metric, proposed by
+
+Exploring CLIP for Assessing the Look and Feel of Images.
+Jianyi Wang Kelvin C.K. Chan Chen Change Loy.
+AAAI 2023.
+
+Ref url: https://github.com/IceClear/CLIP-IQA
+Re-implmented by: Chaofeng Chen (https://github.com/chaofengc) with the following modification:
+    - We assemble multiple prompts to improve the results of clipiqa model.
+
+"""
+import torch
+import torch.nn as nn
+import sys
+
+import pyiqa
+from pyiqa.archs.arch_util import load_file_from_url
+from pyiqa.archs.arch_util import load_pretrained_network
+
+import clip
+from .constants import OPENAI_CLIP_MEAN, OPENAI_CLIP_STD
+from .clip_model import load
+
+
+default_model_urls = {
+    'clipiqa+': 'https://github.com/chaofengc/IQA-PyTorch/releases/download/v0.1-weights/CLIP-IQA+_learned_prompts-603f3273.pth',
+    'clipiqa+_rn50_512': 'https://github.com/chaofengc/IQA-PyTorch/releases/download/v0.1-weights/CLIPIQA+_RN50_512-89f5d940.pth',
+    'clipiqa+_vitL14_512': 'https://github.com/chaofengc/IQA-PyTorch/releases/download/v0.1-weights/CLIPIQA+_ViTL14_512-e66488f2.pth',
+}
+
+
+class PromptLearner(nn.Module):
+    """
+    Disclaimer:
+        This implementation follows exactly the official codes in: https://github.com/IceClear/CLIP-IQA. We have no idea why some tricks are implemented like this, which include
+            1. Using n_ctx prefix characters "X"
+            2. Appending extra "." at the end
+            3. Insert the original text embedding at the middle
+    """
+
+    def __init__(self, clip_model, n_ctx=16) -> None:
+        super().__init__()
+
+        # For the following codes about prompts, we follow the official codes to get the same results
+        prompt_prefix = " ".join(["X"] * n_ctx) + ' '
+        init_prompts = [prompt_prefix + 'Good photo..', prompt_prefix + 'Bad photo..']
+        with torch.no_grad():
+            txt_token = clip.tokenize(init_prompts)
+            self.tokenized_prompts = txt_token
+            init_embedding = clip_model.token_embedding(txt_token)
+
+        init_ctx = init_embedding[:, 1: 1 + n_ctx]
+        self.ctx = nn.Parameter(init_ctx)
+
+        self.n_ctx = n_ctx
+
+        self.n_cls = len(init_prompts)
+        self.name_lens = [3, 3]  # hard coded length, which does not include the extra "." at the end
+
+        self.register_buffer("token_prefix", init_embedding[:, :1, :])  # SOS
+        self.register_buffer("token_suffix", init_embedding[:, 1 + n_ctx:, :])  # CLS, EOS
+
+    def get_prompts_with_middel_class(self,):
+
+        ctx = self.ctx.to(self.token_prefix)
+        if ctx.dim() == 2:
+            ctx = ctx.unsqueeze(0).expand(self.n_cls, -1, -1)
+
+        half_n_ctx = self.n_ctx // 2
+        prompts = []
+        for i in range(self.n_cls):
+            name_len = self.name_lens[i]
+            prefix_i = self.token_prefix[i: i + 1, :, :]
+            class_i = self.token_suffix[i: i + 1, :name_len, :]
+            suffix_i = self.token_suffix[i: i + 1, name_len:, :]
+            ctx_i_half1 = ctx[i: i + 1, :half_n_ctx, :]
+            ctx_i_half2 = ctx[i: i + 1, half_n_ctx:, :]
+            prompt = torch.cat(
+                [
+                    prefix_i,     # (1, 1, dim)
+                    ctx_i_half1,  # (1, n_ctx//2, dim)
+                    class_i,      # (1, name_len, dim)
+                    ctx_i_half2,  # (1, n_ctx//2, dim)
+                    suffix_i,     # (1, *, dim)
+                ],
+                dim=1,
+            )
+            prompts.append(prompt)
+        prompts = torch.cat(prompts, dim=0)
+        return prompts
+
+    def forward(self, clip_model):
+        prompts = self.get_prompts_with_middel_class()
+        # self.get_prompts_with_middel_class
+        x = prompts + clip_model.positional_embedding.type(clip_model.dtype)
+        x = x.permute(1, 0, 2)  # NLD -> LND
+        x = clip_model.transformer(x)
+        x = x.permute(1, 0, 2)  # LND -> NLD
+        x = clip_model.ln_final(x).type(clip_model.dtype)
+
+        # x.shape = [batch_size, n_ctx, transformer.width]
+        # take features from the eot embedding (eot_token is the highest number in each sequence)
+        x = x[torch.arange(x.shape[0]), self.tokenized_prompts.argmax(dim=-1)] @ clip_model.text_projection
+
+        return x
+
+
+class CLIPIQA(nn.Module):
+    def __init__(self,
+                 model_type='clipiqa+_vitL14_512',
+                 backbone='ViT-L/14',
+                 pretrained=True,
+                 pos_embedding=False,
+                 ) -> None:
+        super().__init__()
+
+        self.clip_model = [load(backbone, 'cpu')]  # avoid saving clip weights
+        # Different from original paper, we assemble multiple prompts to improve performance
+        self.prompt_pairs = clip.tokenize([
+            'Good image', 'bad image',
+            'Sharp image', 'blurry image',
+            'sharp edges', 'blurry edges',
+            'High resolution image', 'low resolution image',
+            'Noise-free image', 'noisy image',
+        ])
+
+        self.model_type = model_type
+        self.pos_embedding = pos_embedding
+        if 'clipiqa+' in model_type:
+            self.prompt_learner = PromptLearner(self.clip_model[0])
+
+        self.default_mean = torch.Tensor(OPENAI_CLIP_MEAN).view(1, 3, 1, 1)
+        self.default_std = torch.Tensor(OPENAI_CLIP_STD).view(1, 3, 1, 1)
+
+        for p in self.clip_model[0].parameters():
+            p.requires_grad = False
+        
+        if pretrained and 'clipiqa+' in model_type:
+            if model_type == 'clipiqa+' and backbone == 'RN50':
+                self.prompt_learner.ctx.data = torch.load(load_file_from_url(default_model_urls['clipiqa+']))
+            elif model_type in default_model_urls.keys():
+                load_pretrained_network(self, default_model_urls[model_type], True, 'params')
+            else:
+                raise(f'No pretrained model for {model_type}')
+    
+
+    def forward(self, x, multi=False, layer=-1):
+        # no need to preprocess image here
+        # as already image is already preprocessed
+        # x = (x - self.default_mean.to(x)) / self.default_std.to(x)
+        clip_model = self.clip_model[0].to(x)
+
+        if self.model_type == 'clipiqa':
+            prompts = self.prompt_pairs.to(x.device)
+            logits_per_image, logits_per_text, image_feature, token_feature = clip_model(x, prompts, pos_embedding=self.pos_embedding)
+        elif 'clipiqa+' in self.model_type:
+            # learned_prompt_feature = self.prompt_learner(clip_model)
+            learned_prompt_feature = 0
+            logits_per_image, logits_per_text, image_feature, token_feature = clip_model(
+                x, None, text_features=learned_prompt_feature,  pos_embedding=self.pos_embedding)
+
+        # probs = logits_per_image.reshape(logits_per_image.shape[0], -1, 2).softmax(dim=-1)
+
+        # return probs[..., 0].mean(dim=1, keepdim=True), image_feature
+        return image_feature, token_feature

cover/models/constants.py

@@ -0,0 +1,8 @@
+IMAGENET_DEFAULT_MEAN = (0.485, 0.456, 0.406)
+IMAGENET_DEFAULT_STD = (0.229, 0.224, 0.225)
+IMAGENET_INCEPTION_MEAN = (0.5, 0.5, 0.5)
+IMAGENET_INCEPTION_STD = (0.5, 0.5, 0.5)
+IMAGENET_DPN_MEAN = (124 / 255, 117 / 255, 104 / 255)
+IMAGENET_DPN_STD = tuple([1 / (.0167 * 255)] * 3)
+OPENAI_CLIP_MEAN = (122.77, 116.75, 104.09)
+OPENAI_CLIP_STD = (68.50, 66.63, 70.32)

cover/models/conv_backbone.py

@@ -0,0 +1,651 @@
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+from timm.models.layers import trunc_normal_, DropPath
+from timm.models.registry import register_model
+from .clipiqa_arch import CLIPIQA
+
+
+class GRN(nn.Module):
+    """ GRN (Global Response Normalization) layer
+    """
+    def __init__(self, dim):
+        super().__init__()
+        self.gamma = nn.Parameter(torch.zeros(1, 1, 1, dim))
+        self.beta = nn.Parameter(torch.zeros(1, 1, 1, dim))
+
+    def forward(self, x):
+        Gx = torch.norm(x, p=2, dim=(1,2), keepdim=True)
+        Nx = Gx / (Gx.mean(dim=-1, keepdim=True) + 1e-6)
+        return self.gamma * (x * Nx) + self.beta + x
+
+class Block(nn.Module):
+    r""" ConvNeXt Block. There are two equivalent implementations:
+    (1) DwConv -> LayerNorm (channels_first) -> 1x1 Conv -> GELU -> 1x1 Conv; all in (N, C, H, W)
+    (2) DwConv -> Permute to (N, H, W, C); LayerNorm (channels_last) -> Linear -> GELU -> Linear; Permute back
+    We use (2) as we find it slightly faster in PyTorch
+    
+    Args:
+        dim (int): Number of input channels.
+        drop_path (float): Stochastic depth rate. Default: 0.0
+        layer_scale_init_value (float): Init value for Layer Scale. Default: 1e-6.
+    """
+    def __init__(self, dim, drop_path=0., layer_scale_init_value=1e-6):
+        super().__init__()
+        self.dwconv = nn.Conv2d(dim, dim, kernel_size=7, padding=3, groups=dim) # depthwise conv
+        self.norm = LayerNorm(dim, eps=1e-6)
+        self.pwconv1 = nn.Linear(dim, 4 * dim) # pointwise/1x1 convs, implemented with linear layers
+        self.act = nn.GELU()
+        self.pwconv2 = nn.Linear(4 * dim, dim)
+        self.gamma = nn.Parameter(layer_scale_init_value * torch.ones((dim)), 
+                                    requires_grad=True) if layer_scale_init_value > 0 else None
+        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
+
+    def forward(self, x):
+        input = x
+        x = self.dwconv(x)
+        x = x.permute(0, 2, 3, 1) # (N, C, H, W) -> (N, H, W, C)
+        x = self.norm(x)
+        x = self.pwconv1(x)
+        x = self.act(x)
+        x = self.pwconv2(x)
+        if self.gamma is not None:
+            x = self.gamma * x
+        x = x.permute(0, 3, 1, 2) # (N, H, W, C) -> (N, C, H, W)
+
+        x = input + self.drop_path(x)
+        return x
+
+class ConvNeXt(nn.Module):
+    r""" ConvNeXt
+        A PyTorch impl of : `A ConvNet for the 2020s`  -
+          https://arxiv.org/pdf/2201.03545.pdf
+    Args:
+        in_chans (int): Number of input image channels. Default: 3
+        num_classes (int): Number of classes for classification head. Default: 1000
+        depths (tuple(int)): Number of blocks at each stage. Default: [3, 3, 9, 3]
+        dims (int): Feature dimension at each stage. Default: [96, 192, 384, 768]
+        drop_path_rate (float): Stochastic depth rate. Default: 0.
+        layer_scale_init_value (float): Init value for Layer Scale. Default: 1e-6.
+        head_init_scale (float): Init scaling value for classifier weights and biases. Default: 1.
+    """
+    def __init__(self, in_chans=3, num_classes=1000, 
+                 depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], drop_path_rate=0., 
+                 layer_scale_init_value=1e-6, head_init_scale=1.,
+                 ):
+        super().__init__()
+
+        self.downsample_layers = nn.ModuleList() # stem and 3 intermediate downsampling conv layers
+        stem = nn.Sequential(
+            nn.Conv2d(in_chans, dims[0], kernel_size=4, stride=4),
+            LayerNorm(dims[0], eps=1e-6, data_format="channels_first")
+        )
+        self.downsample_layers.append(stem)
+        for i in range(3):
+            downsample_layer = nn.Sequential(
+                    LayerNorm(dims[i], eps=1e-6, data_format="channels_first"),
+                    nn.Conv2d(dims[i], dims[i+1], kernel_size=2, stride=2),
+            )
+            self.downsample_layers.append(downsample_layer)
+
+        self.stages = nn.ModuleList() # 4 feature resolution stages, each consisting of multiple residual blocks
+        dp_rates=[x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] 
+        cur = 0
+        for i in range(4):
+            stage = nn.Sequential(
+                *[Block(dim=dims[i], drop_path=dp_rates[cur + j], 
+                layer_scale_init_value=layer_scale_init_value) for j in range(depths[i])]
+            )
+            self.stages.append(stage)
+            cur += depths[i]
+
+        self.norm = nn.LayerNorm(dims[-1], eps=1e-6) # final norm layer
+        self.head = nn.Linear(dims[-1], num_classes)
+
+        self.apply(self._init_weights)
+        self.head.weight.data.mul_(head_init_scale)
+        self.head.bias.data.mul_(head_init_scale)
+
+    def _init_weights(self, m):
+        if isinstance(m, (nn.Conv2d, nn.Linear)):
+            trunc_normal_(m.weight, std=.02)
+            nn.init.constant_(m.bias, 0)
+
+    def forward_features(self, x):
+        for i in range(4):
+            x = self.downsample_layers[i](x)
+            x = self.stages[i](x)
+        return self.norm(x.mean([-2, -1])) # global average pooling, (N, C, H, W) -> (N, C)
+
+    def forward(self, x):
+        x = self.forward_features(x)
+        x = self.head(x)
+        return x
+
+class LayerNorm(nn.Module):
+    r""" LayerNorm that supports two data formats: channels_last (default) or channels_first. 
+    The ordering of the dimensions in the inputs. channels_last corresponds to inputs with 
+    shape (batch_size, height, width, channels) while channels_first corresponds to inputs 
+    with shape (batch_size, channels, height, width).
+    """
+    def __init__(self, normalized_shape, eps=1e-6, data_format="channels_last"):
+        super().__init__()
+        self.weight = nn.Parameter(torch.ones(normalized_shape))
+        self.bias = nn.Parameter(torch.zeros(normalized_shape))
+        self.eps = eps
+        self.data_format = data_format
+        if self.data_format not in ["channels_last", "channels_first"]:
+            raise NotImplementedError 
+        self.normalized_shape = (normalized_shape, )
+    
+    def forward(self, x):
+        if self.data_format == "channels_last":
+            return F.layer_norm(x, self.normalized_shape, self.weight, self.bias, self.eps)
+        elif self.data_format == "channels_first":
+            u = x.mean(1, keepdim=True)
+            s = (x - u).pow(2).mean(1, keepdim=True)
+            x = (x - u) / torch.sqrt(s + self.eps)
+            if len(x.shape) == 4:
+                x = self.weight[:, None, None] * x + self.bias[:, None, None]
+            elif len(x.shape) == 5:
+                x = self.weight[:, None, None, None] * x + self.bias[:, None, None, None]
+            return x
+        
+        
+class Block3D(nn.Module):
+    r""" ConvNeXt Block. There are two equivalent implementations:
+    (1) DwConv -> LayerNorm (channels_first) -> 1x1 Conv -> GELU -> 1x1 Conv; all in (N, C, H, W)
+    (2) DwConv -> Permute to (N, H, W, C); LayerNorm (channels_last) -> Linear -> GELU -> Linear; Permute back
+    We use (2) as we find it slightly faster in PyTorch
+    
+    Args:
+        dim (int): Number of input channels.
+        drop_path (float): Stochastic depth rate. Default: 0.0
+        layer_scale_init_value (float): Init value for Layer Scale. Default: 1e-6.
+    """
+    def __init__(self, dim, drop_path=0., inflate_len=3, layer_scale_init_value=1e-6):
+        super().__init__()
+        self.dwconv = nn.Conv3d(dim, dim, kernel_size=(inflate_len,7,7), padding=(inflate_len // 2,3,3), groups=dim) # depthwise conv
+        self.norm = LayerNorm(dim, eps=1e-6)
+        self.pwconv1 = nn.Linear(dim, 4 * dim) # pointwise/1x1 convs, implemented with linear layers
+        self.act = nn.GELU()
+        self.pwconv2 = nn.Linear(4 * dim, dim)
+        self.gamma = nn.Parameter(layer_scale_init_value * torch.ones((dim)), 
+                                    requires_grad=True) if layer_scale_init_value > 0 else None
+        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
+
+    def forward(self, x):
+        input = x
+        x = self.dwconv(x)
+        x = x.permute(0, 2, 3, 4, 1) # (N, C, H, W) -> (N, H, W, C)
+        x = self.norm(x)
+        x = self.pwconv1(x)
+        x = self.act(x)
+        x = self.pwconv2(x)
+        if self.gamma is not None:
+            x = self.gamma * x
+        x = x.permute(0, 4, 1, 2, 3) # (N, H, W, C) -> (N, C, H, W)
+
+        x = input + self.drop_path(x)
+        return x
+    
+class BlockV2(nn.Module):
+    """ ConvNeXtV2 Block.
+    
+    Args:
+        dim (int): Number of input channels.
+        drop_path (float): Stochastic depth rate. Default: 0.0
+    """
+    def __init__(self, dim, drop_path=0.):
+        super().__init__()
+        self.dwconv = nn.Conv2d(dim, dim, kernel_size=7, padding=3, groups=dim) # depthwise conv
+        self.norm = LayerNorm(dim, eps=1e-6)
+        self.pwconv1 = nn.Linear(dim, 4 * dim) # pointwise/1x1 convs, implemented with linear layers
+        self.act = nn.GELU()
+        self.grn = GRN(4 * dim)
+        self.pwconv2 = nn.Linear(4 * dim, dim)
+        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
+
+    def forward(self, x):
+        input = x
+        x = self.dwconv(x)
+        x = x.permute(0, 2, 3, 1) # (N, C, H, W) -> (N, H, W, C)
+        x = self.norm(x)
+        x = self.pwconv1(x)
+        x = self.act(x)
+        x = self.grn(x)
+        x = self.pwconv2(x)
+        x = x.permute(0, 3, 1, 2) # (N, H, W, C) -> (N, C, H, W)
+
+        x = input + self.drop_path(x)
+        return x
+    
+class BlockV23D(nn.Module):
+    """ ConvNeXtV2 Block.
+    
+    Args:
+        dim (int): Number of input channels.
+        drop_path (float): Stochastic depth rate. Default: 0.0
+    """
+    def __init__(self, dim, drop_path=0., inflate_len=3,):
+        super().__init__()
+        self.dwconv = nn.Conv3d(dim, dim, kernel_size=(inflate_len,7,7), padding=(inflate_len // 2,3,3), groups=dim) # depthwise conv
+        self.norm = LayerNorm(dim, eps=1e-6)
+        self.pwconv1 = nn.Linear(dim, 4 * dim) # pointwise/1x1 convs, implemented with linear layers
+        self.act = nn.GELU()
+        self.grn = GRN(4 * dim)
+        self.pwconv2 = nn.Linear(4 * dim, dim)
+        self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity()
+
+    def forward(self, x):
+        input = x
+        x = self.dwconv(x)
+        x = x.permute(0, 2, 3, 4, 1) # (N, C, H, W) -> (N, H, W, C)
+        x = self.norm(x)
+        x = self.pwconv1(x)
+        x = self.act(x)
+        x = self.grn(x)
+        x = self.pwconv2(x)
+        x = x.permute(0, 4, 1, 2, 3) # (N, H, W, C) -> (N, C, H, W)
+
+        x = input + self.drop_path(x)
+        return x
+
+class ConvNeXtV2(nn.Module):
+    """ ConvNeXt V2
+        
+    Args:
+        in_chans (int): Number of input image channels. Default: 3
+        num_classes (int): Number of classes for classification head. Default: 1000
+        depths (tuple(int)): Number of blocks at each stage. Default: [3, 3, 9, 3]
+        dims (int): Feature dimension at each stage. Default: [96, 192, 384, 768]
+        drop_path_rate (float): Stochastic depth rate. Default: 0.
+        head_init_scale (float): Init scaling value for classifier weights and biases. Default: 1.
+    """
+    def __init__(self, in_chans=3, num_classes=1000, 
+                 depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], 
+                 drop_path_rate=0., head_init_scale=1.
+                 ):
+        super().__init__()
+        self.depths = depths
+        self.downsample_layers = nn.ModuleList() # stem and 3 intermediate downsampling conv layers
+        stem = nn.Sequential(
+            nn.Conv2d(in_chans, dims[0], kernel_size=4, stride=4),
+            LayerNorm(dims[0], eps=1e-6, data_format="channels_first")
+        )
+        self.downsample_layers.append(stem)
+        for i in range(3):
+            downsample_layer = nn.Sequential(
+                    LayerNorm(dims[i], eps=1e-6, data_format="channels_first"),
+                    nn.Conv2d(dims[i], dims[i+1], kernel_size=2, stride=2),
+            )
+            self.downsample_layers.append(downsample_layer)
+
+        self.stages = nn.ModuleList() # 4 feature resolution stages, each consisting of multiple residual blocks
+        dp_rates=[x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] 
+        cur = 0
+        for i in range(4):
+            stage = nn.Sequential(
+                *[BlockV2(dim=dims[i], drop_path=dp_rates[cur + j]) for j in range(depths[i])]
+            )
+            self.stages.append(stage)
+            cur += depths[i]
+
+        self.norm = nn.LayerNorm(dims[-1], eps=1e-6) # final norm layer
+        self.head = nn.Linear(dims[-1], num_classes)
+
+        self.apply(self._init_weights)
+        self.head.weight.data.mul_(head_init_scale)
+        self.head.bias.data.mul_(head_init_scale)
+
+    def _init_weights(self, m):
+        if isinstance(m, (nn.Conv2d, nn.Linear)):
+            trunc_normal_(m.weight, std=.02)
+            nn.init.constant_(m.bias, 0)
+
+    def forward_features(self, x):
+        for i in range(4):
+            x = self.downsample_layers[i](x)
+            x = self.stages[i](x)
+        return self.norm(x.mean([-2, -1])) # global average pooling, (N, C, H, W) -> (N, C)
+
+    def forward(self, x):
+        x = self.forward_features(x)
+        x = self.head(x)
+        return x
+
+def convnextv2_atto(**kwargs):
+    model = ConvNeXtV2(depths=[2, 2, 6, 2], dims=[40, 80, 160, 320], **kwargs)
+    return model
+
+def convnextv2_femto(**kwargs):
+    model = ConvNeXtV2(depths=[2, 2, 6, 2], dims=[48, 96, 192, 384], **kwargs)
+    return model
+
+def convnext_pico(**kwargs):
+    model = ConvNeXtV2(depths=[2, 2, 6, 2], dims=[64, 128, 256, 512], **kwargs)
+    return model
+
+def convnextv2_nano(**kwargs):
+    model = ConvNeXtV2(depths=[2, 2, 8, 2], dims=[80, 160, 320, 640], **kwargs)
+    return model
+
+def convnextv2_tiny(**kwargs):
+    model = ConvNeXtV2(depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], **kwargs)
+    return model
+
+def convnextv2_base(**kwargs):
+    model = ConvNeXtV2(depths=[3, 3, 27, 3], dims=[128, 256, 512, 1024], **kwargs)
+    return model
+
+def convnextv2_large(**kwargs):
+    model = ConvNeXtV2(depths=[3, 3, 27, 3], dims=[192, 384, 768, 1536], **kwargs)
+    return model
+
+def convnextv2_huge(**kwargs):
+    model = ConvNeXtV2(depths=[3, 3, 27, 3], dims=[352, 704, 1408, 2816], **kwargs)
+    return model
+    
+class ConvNeXt3D(nn.Module):
+    r""" ConvNeXt
+        A PyTorch impl of : `A ConvNet for the 2020s`  -
+          https://arxiv.org/pdf/2201.03545.pdf
+    Args:
+        in_chans (int): Number of input image channels. Default: 3
+        num_classes (int): Number of classes for classification head. Default: 1000
+        depths (tuple(int)): Number of blocks at each stage. Default: [3, 3, 9, 3]
+        dims (int): Feature dimension at each stage. Default: [96, 192, 384, 768]
+        drop_path_rate (float): Stochastic depth rate. Default: 0.
+        layer_scale_init_value (float): Init value for Layer Scale. Default: 1e-6.
+        head_init_scale (float): Init scaling value for classifier weights and biases. Default: 1.
+    """
+    def __init__(self, in_chans=3, num_classes=1000,
+                 inflate_strategy='131',
+                 depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], drop_path_rate=0., 
+                 layer_scale_init_value=1e-6, head_init_scale=1.,
+                 ):
+        super().__init__()
+
+        self.downsample_layers = nn.ModuleList() # stem and 3 intermediate downsampling conv layers
+        stem = nn.Sequential(
+            nn.Conv3d(in_chans, dims[0], kernel_size=(2,4,4), stride=(2,4,4)),
+            LayerNorm(dims[0], eps=1e-6, data_format="channels_first")
+        )
+        self.downsample_layers.append(stem)
+        for i in range(3):
+            downsample_layer = nn.Sequential(
+                    LayerNorm(dims[i], eps=1e-6, data_format="channels_first"),
+                    nn.Conv3d(dims[i], dims[i+1], kernel_size=(1,2,2), stride=(1,2,2)),
+            )
+            self.downsample_layers.append(downsample_layer)
+
+        self.stages = nn.ModuleList() # 4 feature resolution stages, each consisting of multiple residual blocks
+        dp_rates=[x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] 
+        cur = 0
+        for i in range(4):
+            stage = nn.Sequential(
+                *[Block3D(dim=dims[i], inflate_len=int(inflate_strategy[j%len(inflate_strategy)]),
+                          drop_path=dp_rates[cur + j], 
+                layer_scale_init_value=layer_scale_init_value) for j in range(depths[i])]
+            )
+            self.stages.append(stage)
+            cur += depths[i]
+
+        self.norm = nn.LayerNorm(dims[-1], eps=1e-6) # final norm layer
+
+        self.apply(self._init_weights)
+        
+    def inflate_weights(self, s_state_dict):
+        t_state_dict = self.state_dict()
+        from collections import OrderedDict
+        for key in t_state_dict.keys():
+            if key not in s_state_dict:
+                print(key)
+                continue
+            if t_state_dict[key].shape != s_state_dict[key].shape:
+                t = t_state_dict[key].shape[2]
+                s_state_dict[key] = s_state_dict[key].unsqueeze(2).repeat(1,1,t,1,1) / t
+        self.load_state_dict(s_state_dict, strict=False)
+
+    def _init_weights(self, m):
+        if isinstance(m, (nn.Conv3d, nn.Linear)):
+            trunc_normal_(m.weight, std=.02)
+            nn.init.constant_(m.bias, 0)
+
+    def forward_features(self, x, return_spatial=False, multi=False, layer=-1):
+        if multi:
+            xs = []
+        for i in range(4):
+            x = self.downsample_layers[i](x)
+            x = self.stages[i](x)
+            if multi:
+                xs.append(x)
+        if return_spatial:
+            if multi:
+                shape = xs[-1].shape[2:]
+                return torch.cat([F.interpolate(x,size=shape, mode="trilinear") for x in xs[:-1]], 1) #+ [self.norm(x.permute(0, 2, 3, 4, 1)).permute(0, 4, 1, 2, 3)], 1)
+            elif layer > -1:
+                return xs[layer]
+            else:
+                return self.norm(x.permute(0, 2, 3, 4, 1)).permute(0, 4, 1, 2, 3)
+        return self.norm(x.mean([-3, -2, -1])) # global average pooling, (N, C, T, H, W) -> (N, C)
+
+    def forward(self, x, multi=False, layer=-1):
+        x = self.forward_features(x, True, multi=multi, layer=layer)
+        return x
+
+
+class ConvNeXtV23D(nn.Module):
+    """ ConvNeXt V2
+        
+    Args:
+        in_chans (int): Number of input image channels. Default: 3
+        num_classes (int): Number of classes for classification head. Default: 1000
+        depths (tuple(int)): Number of blocks at each stage. Default: [3, 3, 9, 3]
+        dims (int): Feature dimension at each stage. Default: [96, 192, 384, 768]
+        drop_path_rate (float): Stochastic depth rate. Default: 0.
+        head_init_scale (float): Init scaling value for classifier weights and biases. Default: 1.
+    """
+    def __init__(self, in_chans=3, num_classes=1000, 
+                 inflate_strategy='131',
+                 depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], 
+                 drop_path_rate=0., head_init_scale=1.
+                 ):
+        super().__init__()
+        self.depths = depths
+        self.downsample_layers = nn.ModuleList() # stem and 3 intermediate downsampling conv layers
+        stem = nn.Sequential(
+            nn.Conv3d(in_chans, dims[0], kernel_size=(2,4,4), stride=(2,4,4)),
+            LayerNorm(dims[0], eps=1e-6, data_format="channels_first")
+        )
+        self.downsample_layers.append(stem)
+        for i in range(3):
+            downsample_layer = nn.Sequential(
+                    LayerNorm(dims[i], eps=1e-6, data_format="channels_first"),
+                    nn.Conv3d(dims[i], dims[i+1], kernel_size=(1,2,2), stride=(1,2,2)),
+            )
+            self.downsample_layers.append(downsample_layer)
+
+        self.stages = nn.ModuleList() # 4 feature resolution stages, each consisting of multiple residual blocks
+        dp_rates=[x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] 
+        cur = 0
+        for i in range(4):
+            stage = nn.Sequential(
+                *[BlockV23D(dim=dims[i], drop_path=dp_rates[cur + j],
+                            inflate_len=int(inflate_strategy[j%len(inflate_strategy)]),
+                           ) for j in range(depths[i])]
+            )
+            self.stages.append(stage)
+            cur += depths[i]
+
+        self.norm = nn.LayerNorm(dims[-1], eps=1e-6) # final norm layer
+        self.head = nn.Linear(dims[-1], num_classes)
+
+        self.apply(self._init_weights)
+        self.head.weight.data.mul_(head_init_scale)
+        self.head.bias.data.mul_(head_init_scale)
+    
+    def inflate_weights(self, pretrained_path):
+        t_state_dict = self.state_dict()
+        s_state_dict = torch.load(pretrained_path)["model"]
+        from collections import OrderedDict
+        for key in t_state_dict.keys():
+            if key not in s_state_dict:
+                print(key)
+                continue
+            if t_state_dict[key].shape != s_state_dict[key].shape:
+                print(t_state_dict[key].shape, s_state_dict[key].shape)
+                t = t_state_dict[key].shape[2]
+                s_state_dict[key] = s_state_dict[key].unsqueeze(2).repeat(1,1,t,1,1) / t
+        self.load_state_dict(s_state_dict, strict=False)
+
+    def _init_weights(self, m):
+        if isinstance(m, (nn.Conv3d, nn.Linear)):
+            trunc_normal_(m.weight, std=.02)
+            nn.init.constant_(m.bias, 0)
+
+    def forward_features(self, x, return_spatial=False, multi=False, layer=-1):
+        if multi:
+            xs = []
+        for i in range(4):
+            x = self.downsample_layers[i](x)
+            x = self.stages[i](x)
+            if multi:
+                xs.append(x)
+        if return_spatial:
+            if multi:
+                shape = xs[-1].shape[2:]
+                return torch.cat([F.interpolate(x,size=shape, mode="trilinear") for x in xs[:-1]], 1) #+ [self.norm(x.permute(0, 2, 3, 4, 1)).permute(0, 4, 1, 2, 3)], 1)
+            elif layer > -1:
+                return xs[layer]
+            else:
+                return self.norm(x.permute(0, 2, 3, 4, 1)).permute(0, 4, 1, 2, 3)
+        return self.norm(x.mean([-3, -2, -1])) # global average pooling, (N, C, T, H, W) -> (N, C)
+
+    def forward(self, x, multi=False, layer=-1):
+        x = self.forward_features(x, True, multi=multi, layer=layer)
+        return x
+
+
+model_urls = {
+    "convnext_tiny_1k": "https://dl.fbaipublicfiles.com/convnext/convnext_tiny_1k_224_ema.pth",
+    "convnext_small_1k": "https://dl.fbaipublicfiles.com/convnext/convnext_small_1k_224_ema.pth",
+    "convnext_base_1k": "https://dl.fbaipublicfiles.com/convnext/convnext_base_1k_224_ema.pth",
+    "convnext_large_1k": "https://dl.fbaipublicfiles.com/convnext/convnext_large_1k_224_ema.pth",
+    "convnext_tiny_22k": "https://dl.fbaipublicfiles.com/convnext/convnext_tiny_22k_224.pth",
+    "convnext_small_22k": "https://dl.fbaipublicfiles.com/convnext/convnext_small_22k_224.pth",
+    "convnext_base_22k": "https://dl.fbaipublicfiles.com/convnext/convnext_base_22k_224.pth",
+    "convnext_large_22k": "https://dl.fbaipublicfiles.com/convnext/convnext_large_22k_224.pth",
+    "convnext_xlarge_22k": "https://dl.fbaipublicfiles.com/convnext/convnext_xlarge_22k_224.pth",
+}
+
+def convnext_tiny(pretrained=False,in_22k=False, **kwargs):
+    model = ConvNeXt(depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], **kwargs)
+    if pretrained:
+        url = model_urls['convnext_tiny_22k'] if in_22k else model_urls['convnext_tiny_1k']
+        checkpoint = torch.hub.load_state_dict_from_url(url=url, map_location="cpu", check_hash=True)
+        model.load_state_dict(checkpoint["model"])
+    return model
+
+def convnext_small(pretrained=False,in_22k=False, **kwargs):
+    model = ConvNeXt(depths=[3, 3, 27, 3], dims=[96, 192, 384, 768], **kwargs)
+    if pretrained:
+        url = model_urls['convnext_small_22k'] if in_22k else model_urls['convnext_small_1k']
+        checkpoint = torch.hub.load_state_dict_from_url(url=url, map_location="cpu")
+        model.load_state_dict(checkpoint["model"])
+    return model
+
+def convnext_base(pretrained=False, in_22k=False, **kwargs):
+    model = ConvNeXt(depths=[3, 3, 27, 3], dims=[128, 256, 512, 1024], **kwargs)
+    if pretrained:
+        url = model_urls['convnext_base_22k'] if in_22k else model_urls['convnext_base_1k']
+        checkpoint = torch.hub.load_state_dict_from_url(url=url, map_location="cpu")
+        model.load_state_dict(checkpoint["model"])
+    return model
+
+
+def convnext_large(pretrained=False, in_22k=False, **kwargs):
+    model = ConvNeXt(depths=[3, 3, 27, 3], dims=[192, 384, 768, 1536], **kwargs)
+    if pretrained:
+        url = model_urls['convnext_large_22k'] if in_22k else model_urls['convnext_large_1k']
+        checkpoint = torch.hub.load_state_dict_from_url(url=url, map_location="cpu")
+        model.load_state_dict(checkpoint["model"])
+    return model
+
+def convnext_xlarge(pretrained=False, in_22k=False, **kwargs):
+    model = ConvNeXt(depths=[3, 3, 27, 3], dims=[256, 512, 1024, 2048], **kwargs)
+    if pretrained:
+        assert in_22k, "only ImageNet-22K pre-trained ConvNeXt-XL is available; please set in_22k=True"
+        url = model_urls['convnext_xlarge_22k']
+        checkpoint = torch.hub.load_state_dict_from_url(url=url, map_location="cpu")
+        model.load_state_dict(checkpoint["model"])
+        
+    return model
+
+def convnext_3d_tiny(pretrained=False, in_22k=False, **kwargs):
+    print("Using Imagenet 22K pretrain", in_22k)
+    model = ConvNeXt3D(depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], **kwargs)
+    if pretrained:
+        url = model_urls['convnext_tiny_22k'] if in_22k else model_urls['convnext_tiny_1k']
+        checkpoint = torch.hub.load_state_dict_from_url(url=url, map_location="cpu", check_hash=True)
+        model.inflate_weights(checkpoint["model"])
+    return model
+
+def convnext_3d_small(pretrained=False, in_22k=False, **kwargs):
+    model = ConvNeXt3D(depths=[3, 3, 27, 3], dims=[96, 192, 384, 768], **kwargs)
+    if pretrained:
+        url = model_urls['convnext_small_22k'] if in_22k else model_urls['convnext_small_1k']
+        checkpoint = torch.hub.load_state_dict_from_url(url=url, map_location="cpu", check_hash=True)
+        model.inflate_weights(checkpoint["model"])
+        
+    return model
+
+def convnextv2_3d_atto(**kwargs):
+    model = ConvNeXtV23D(depths=[2, 2, 6, 2], dims=[40, 80, 160, 320], **kwargs)
+    
+    return model
+
+def convnextv2_3d_femto(pretrained="../pretrained/convnextv2_femto_1k_224_ema.pt", **kwargs):
+    model = ConvNeXtV23D(depths=[2, 2, 6, 2], dims=[48, 96, 192, 384], **kwargs)
+    #model.inflate_weights(pretrained)
+    return model
+
+def convnextv2_3d_pico(pretrained="../pretrained/convnextv2_pico_1k_224_ema.pt", **kwargs):
+    model = ConvNeXtV23D(depths=[2, 2, 6, 2], dims=[64, 128, 256, 512], **kwargs)
+    #model.inflate_weights(pretrained)
+    return model
+
+def convnextv2_3d_nano(pretrained="../pretrained/convnextv2_nano_1k_224_ema.pt", **kwargs):
+    model = ConvNeXtV23D(depths=[2, 2, 8, 2], dims=[80, 160, 320, 640], **kwargs)
+    #model.inflate_weights(pretrained)
+    return model
+
+def convnextv2_tiny(**kwargs):
+    model = ConvNeXtV23D(depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], **kwargs)
+    return model
+
+def convnextv2_base(**kwargs):
+    model = ConvNeXtV23D(depths=[3, 3, 27, 3], dims=[128, 256, 512, 1024], **kwargs)
+    return model
+
+def convnextv2_large(**kwargs):
+    model = ConvNeXtV23D(depths=[3, 3, 27, 3], dims=[192, 384, 768, 1536], **kwargs)
+    return model
+
+def convnextv2_huge(**kwargs):
+    model = ConvNeXtV2(depths=[3, 3, 27, 3], dims=[352, 704, 1408, 2816], **kwargs)
+    return model
+
+def clip_vitL14(pretrained, **kwargs):
+    model = CLIPIQA(model_type='clipiqa+_vitL14_512', backbone='ViT-L/14', pretrained=pretrained)
+    return model
+        
+if __name__ == "__main__":
+
+    device = "cuda" if torch.cuda.is_available() else "cpu"
+    model = convnext_3d_tiny(True).to(device)
+    print(model)
+    from thop import profile
+    print(profile(model, (torch.randn(4,3,32,224,224).to(device),))[0] / 1e9)
+    
+    

cover/models/evaluator.py

@@ -0,0 +1,374 @@
+import time
+from functools import partial, reduce
+
+import torch
+import torch.nn as nn
+from torch.nn.functional import adaptive_avg_pool3d
+
+from .conv_backbone import convnext_3d_small, convnext_3d_tiny, convnextv2_3d_pico, convnextv2_3d_femto, clip_vitL14
+from .head import IQAHead, VARHead, VQAHead
+from .swin_backbone import SwinTransformer2D as ImageBackbone
+from .swin_backbone import SwinTransformer3D as VideoBackbone
+from .swin_backbone import swin_3d_small, swin_3d_tiny
+
+
+class BaseEvaluator(nn.Module):
+    def __init__(
+        self, backbone=dict(), vqa_head=dict(),
+    ):
+        super().__init__()
+        self.backbone = VideoBackbone(**backbone)
+        self.vqa_head = VQAHead(**vqa_head)
+
+    def forward(self, vclip, inference=True, **kwargs):
+        if inference:
+            self.eval()
+            with torch.no_grad():
+                feat = self.backbone(vclip)
+                score = self.vqa_head(feat)
+            self.train()
+            return score
+        else:
+            feat = self.backbone(vclip)
+            score = self.vqa_head(feat)
+            return score
+
+    def forward_with_attention(self, vclip):
+        self.eval()
+        with torch.no_grad():
+            feat, avg_attns = self.backbone(vclip, require_attn=True)
+            score = self.vqa_head(feat)
+            return score, avg_attns
+
+
+class COVER(nn.Module):
+    def __init__(
+        self,
+        backbone_size="divided",
+        backbone_preserve_keys="fragments,resize",
+        multi=False,
+        layer=-1,
+        backbone=dict(
+            resize={"window_size": (4, 4, 4)}, fragments={"window_size": (4, 4, 4)}
+        ),
+        divide_head=False,
+        vqa_head=dict(in_channels=768),
+        var=False,
+    ):
+        self.backbone_preserve_keys = backbone_preserve_keys.split(",")
+        self.multi = multi
+        self.layer = layer
+        super().__init__()
+        for key, hypers in backbone.items():
+            print(backbone_size)
+            if key not in self.backbone_preserve_keys:
+                continue
+            if backbone_size == "divided":
+                t_backbone_size = hypers["type"]
+            else:
+                t_backbone_size = backbone_size
+            if t_backbone_size == "swin_tiny":
+                b = swin_3d_tiny(**backbone[key])
+            elif t_backbone_size == "swin_tiny_grpb":
+                # to reproduce fast-vqa
+                b = VideoBackbone()
+            elif t_backbone_size == "swin_tiny_grpb_m":
+                # to reproduce fast-vqa-m
+                b = VideoBackbone(window_size=(4, 4, 4), frag_biases=[0, 0, 0, 0])
+            elif t_backbone_size == "swin_small":
+                b = swin_3d_small(**backbone[key])
+            elif t_backbone_size == "conv_tiny":
+                b = convnext_3d_tiny(pretrained=True)
+            elif t_backbone_size == "conv_small":
+                b = convnext_3d_small(pretrained=True)
+            elif t_backbone_size == "conv_femto":
+                b = convnextv2_3d_femto(pretrained=True)
+            elif t_backbone_size == "conv_pico":
+                b = convnextv2_3d_pico(pretrained=True)
+            elif t_backbone_size == "xclip":
+                raise NotImplementedError
+            elif t_backbone_size == "clip_iqa+":
+                b = clip_vitL14(pretrained=True)
+            else:
+                raise NotImplementedError
+            print("Setting backbone:", key + "_backbone")
+            setattr(self, key + "_backbone", b)
+        if divide_head:
+            for key in backbone:
+                pre_pool = False #if key == "technical" else True
+                if key not in self.backbone_preserve_keys:
+                    continue
+                b = VQAHead(pre_pool=pre_pool, **vqa_head)
+                print("Setting head:", key + "_head")
+                setattr(self, key + "_head", b)
+        else:
+            if var:
+                self.vqa_head = VARHead(**vqa_head)
+                print(b)
+            else:
+                self.vqa_head = VQAHead(**vqa_head)
+        self.smtc_gate_tech = CrossGatingBlock(x_features=768, num_channels=768, block_size=1, 
+                              grid_size=1, upsample_y=False, dropout_rate=0.1, use_bias=True, use_global_mlp=False)
+        self.smtc_gate_aesc = CrossGatingBlock(x_features=768, num_channels=768, block_size=1, 
+                              grid_size=1, upsample_y=False, dropout_rate=0.1, use_bias=True, use_global_mlp=False)
+
+    def forward(
+        self,
+        vclips,
+        inference=True,
+        return_pooled_feats=False,
+        return_raw_feats=False,
+        reduce_scores=False,
+        pooled=False,
+        **kwargs
+    ):
+        assert (return_pooled_feats & return_raw_feats) == False, "Please only choose one kind of features to return"
+        if inference:
+            self.eval()
+            with torch.no_grad():
+                scores = []
+                feats = {}
+                for key in vclips:
+                    if key == 'technical' or key == 'aesthetic':
+                        feat = getattr(self, key.split("_")[0] + "_backbone")(
+                            vclips[key], multi=self.multi, layer=self.layer, **kwargs
+                        )
+                        if key == 'technical':
+                            feat_gated = self.smtc_gate_tech(feats['semantic'], feat)
+                        elif key == 'aesthetic':
+                            feat_gated = self.smtc_gate_aesc(feats['semantic'], feat)
+                        if hasattr(self, key.split("_")[0] + "_head"):
+                            scores += [getattr(self, key.split("_")[0] + "_head")(feat_gated)]
+                        else:
+                            scores += [getattr(self, "vqa_head")(feat_gated)]
+                    elif key == 'semantic':
+                        x = vclips[key].squeeze()
+                        x =  x.permute(1,0,2,3)
+                        feat, _ = getattr(self, key.split("_")[0] + "_backbone")(
+                            x, multi=self.multi, layer=self.layer, **kwargs
+                        )
+                        # for image feature from clipiqa+ VIT14
+                        # image feature shape (t, c) -> (16, 768)
+                        feat = feat.permute(1,0).contiguous() # (c, t) -> (768, 16)
+                        feat = feat.unsqueeze(-1).unsqueeze(-1) # (c, t, w, h) -> (768, 16, 1, 1)
+                        feat_expand = feat.expand(-1, -1, 7, 7) # (c, t, w, h) -> (768, 16, 7, 7)
+                        feat_expand = feat_expand.unsqueeze(0) # (b, c, t, w, h) -> (1, 768, 16, 7, 7)
+                        if hasattr(self, key.split("_")[0] + "_head"):
+                            score = getattr(self, key.split("_")[0] + "_head")(feat_expand)
+                        else:
+                            score = getattr(self, "vqa_head")(feat_expand)
+                        scores += [score]
+                        feats[key] = feat_expand
+                if reduce_scores:
+                    if len(scores) > 1:
+                        scores = reduce(lambda x, y: x + y, scores)
+                    else:
+                        scores = scores[0]
+                    if pooled:
+                        scores = torch.mean(scores, (1, 2, 3, 4))
+            self.train()
+            if return_pooled_feats or return_raw_feats:
+                return scores, feats
+            return scores
+        else:
+            self.train()
+            scores = []
+            feats = {}
+            for key in vclips:
+                if key == 'technical' or key == 'aesthetic':
+                    feat = getattr(self, key.split("_")[0] + "_backbone")(
+                        vclips[key], multi=self.multi, layer=self.layer, **kwargs
+                    )
+                    if key == 'technical':
+                        feat_gated = self.smtc_gate_tech(feats['semantic'], feat)
+                    elif key == 'aesthetic':
+                        feat_gated = self.smtc_gate_aesc(feats['semantic'], feat)
+                    if hasattr(self, key.split("_")[0] + "_head"):
+                        scores += [getattr(self, key.split("_")[0] + "_head")(feat_gated)]
+                    else:
+                        scores += [getattr(self, "vqa_head")(feat_gated)]
+                    feats[key] = feat
+                elif key == 'semantic':
+                    scores_semantic_list = []
+                    feats_semantic_list = []
+                    for batch_idx in range(vclips[key].shape[0]):
+                        x = vclips[key][batch_idx].squeeze()
+                        x =  x.permute(1,0,2,3)
+                        feat, _ = getattr(self, key.split("_")[0] + "_backbone")(
+                            x, multi=self.multi, layer=self.layer, **kwargs
+                        )
+                        # for image feature from clipiqa+ VIT14
+                        # image feature shape (t, c) -> (16, 768)
+                        feat = feat.permute(1,0).contiguous() # (c, t) -> (768, 16)
+                        feat = feat.unsqueeze(-1).unsqueeze(-1) # (c, t, w, h) -> (768, 16, 1, 1)
+                        feat_expand = feat.expand(-1, -1, 7, 7) # (c, t, w, h) -> (768, 16, 7, 7)
+                        feats_semantic_list.append(feat_expand)
+                        if hasattr(self, key.split("_")[0] + "_head"):
+                            feat_expand = feat_expand.unsqueeze(0) # (b, c, t, w, h) -> (1, 768, 16, 7, 7)
+                            score = getattr(self, key.split("_")[0] + "_head")(feat_expand)
+                            score = score.squeeze(0)
+                            scores_semantic_list.append(score)
+                        else:
+                            feat_expand = feat_expand.unsqueeze(0) # (b, c, t, w, h) -> (1, 768, 16, 7, 7)
+                            score = getattr(self, "vqa_head")(feat_expand)
+                            score = score.squeeze(0)
+                            scores_semantic_list.append(score)
+                    scores_semantic_tensor = torch.stack(scores_semantic_list)
+                    feats[key] = torch.stack(feats_semantic_list)
+                    scores += [scores_semantic_tensor]
+                if return_pooled_feats:
+                    feats[key] = feat.mean((-3, -2, -1))
+            if reduce_scores:
+                if len(scores) > 1:
+                    scores = reduce(lambda x, y: x + y, scores)
+                else:
+                    scores = scores[0]
+                if pooled:
+                    print(scores.shape)
+                    scores = torch.mean(scores, (1, 2, 3, 4))
+                    print(scores.shape)
+
+            if return_pooled_feats:
+                return scores, feats
+            return scores
+
+    def forward_head(
+        self,
+        feats,
+        inference=True,
+        reduce_scores=False,
+        pooled=False,
+        **kwargs
+    ):
+        if inference:
+            self.eval()
+            with torch.no_grad():
+                scores = []
+                feats = {}
+                for key in feats:
+                    feat = feats[key]
+                    if hasattr(self, key.split("_")[0] + "_head"):
+                        scores += [getattr(self, key.split("_")[0] + "_head")(feat)]
+                    else:
+                        scores += [getattr(self, "vqa_head")(feat)]
+                if reduce_scores:
+                    if len(scores) > 1:
+                        scores = reduce(lambda x, y: x + y, scores)
+                    else:
+                        scores = scores[0]
+                    if pooled:
+                        scores = torch.mean(scores, (1, 2, 3, 4))
+            self.train()
+            return scores
+        else:
+            self.train()
+            scores = []
+            feats = {}
+            for key in vclips:
+                feat = getattr(self, key.split("_")[0] + "_backbone")(
+                    vclips[key], multi=self.multi, layer=self.layer, **kwargs
+                )
+                if hasattr(self, key.split("_")[0] + "_head"):
+                    scores += [getattr(self, key.split("_")[0] + "_head")(feat)]
+                else:
+                    scores += [getattr(self, "vqa_head")(feat)]
+                if return_pooled_feats:
+                    feats[key] = feat
+            if reduce_scores:
+                if len(scores) > 1:
+                    scores = reduce(lambda x, y: x + y, scores)
+                else:
+                    scores = scores[0]
+                if pooled:
+                    print(scores.shape)
+                    scores = torch.mean(scores, (1, 2, 3, 4))
+                    print(scores.shape)
+
+            if return_pooled_feats:
+                return scores, feats
+            return scores
+        
+class MinimumCOVER(nn.Module):
+    def __init__(self):
+        super().__init__()
+        self.technical_backbone = VideoBackbone()
+        self.aesthetic_backbone = convnext_3d_tiny(pretrained=True)
+        self.technical_head = VQAHead(pre_pool=False, in_channels=768)
+        self.aesthetic_head = VQAHead(pre_pool=False, in_channels=768)
+
+
+    def forward(self,aesthetic_view, technical_view):
+        self.eval()
+        with torch.no_grad():
+            aesthetic_score = self.aesthetic_head(self.aesthetic_backbone(aesthetic_view))
+            technical_score = self.technical_head(self.technical_backbone(technical_view))
+            
+        aesthetic_score_pooled = torch.mean(aesthetic_score, (1,2,3,4))
+        technical_score_pooled = torch.mean(technical_score, (1,2,3,4))
+        return [aesthetic_score_pooled, technical_score_pooled]
+
+
+
+class BaseImageEvaluator(nn.Module):
+    def __init__(
+        self, backbone=dict(), iqa_head=dict(),
+    ):
+        super().__init__()
+        self.backbone = ImageBackbone(**backbone)
+        self.iqa_head = IQAHead(**iqa_head)
+
+    def forward(self, image, inference=True, **kwargs):
+        if inference:
+            self.eval()
+            with torch.no_grad():
+                feat = self.backbone(image)
+                score = self.iqa_head(feat)
+            self.train()
+            return score
+        else:
+            feat = self.backbone(image)
+            score = self.iqa_head(feat)
+            return score
+
+    def forward_with_attention(self, image):
+        self.eval()
+        with torch.no_grad():
+            feat, avg_attns = self.backbone(image, require_attn=True)
+            score = self.iqa_head(feat)
+            return score, avg_attns
+
+class CrossGatingBlock(nn.Module):  #input shape: n, c, h, w
+    """Cross-gating MLP block."""
+    def __init__(self, x_features, num_channels, block_size, grid_size, cin_y=0,upsample_y=True, use_bias=True, use_global_mlp=True, dropout_rate=0):
+        super().__init__()
+        self.cin_y = cin_y
+        self.x_features = x_features
+        self.num_channels = num_channels
+        self.block_size = block_size
+        self.grid_size = grid_size
+        self.upsample_y = upsample_y
+        self.use_bias = use_bias
+        self.use_global_mlp = use_global_mlp
+        self.drop = dropout_rate
+        self.Conv_0 = nn.Linear(self.x_features, self.num_channels)
+        self.Conv_1 = nn.Linear(self.num_channels, self.num_channels)
+        self.in_project_x = nn.Linear(self.num_channels, self.num_channels, bias=self.use_bias)
+        self.gelu1 = nn.GELU(approximate='tanh')
+        self.out_project_y = nn.Linear(self.num_channels, self.num_channels, bias=self.use_bias)
+        self.dropout1 = nn.Dropout(self.drop)
+    def forward(self, x,y):     #n,c,t,h,w
+        # Upscale Y signal, y is the gating signal.
+        assert y.shape == x.shape
+        x = x.permute(0,2,3,4,1).contiguous()  #n,t,h,w,c
+        y = y.permute(0,2,3,4,1).contiguous()  #n,t,h,w,c
+        x = self.Conv_0(x)
+        y = self.Conv_1(y)
+        shortcut_y = y
+        x = self.in_project_x(x)
+        gx = self.gelu1(x)
+        # Apply cross gating
+        y = y * gx  # gating y using x
+        y = self.out_project_y(y)
+        y = self.dropout1(y)
+        y = y + shortcut_y # y = y * x + y
+        return y.permute(0,4,1,2,3).contiguous()  #n,c,t,h,w

cover/models/head.py

@@ -0,0 +1,101 @@
+import math
+
+import numpy as np
+import torch
+import torch.nn as nn
+from torch.nn import functional as F
+from torchvision.ops import roi_align, roi_pool
+
+
+class VQAHead(nn.Module):
+    """MLP Regression Head for VQA.
+    Args:
+        in_channels: input channels for MLP
+        hidden_channels: hidden channels for MLP
+        dropout_ratio: the dropout ratio for features before the MLP (default 0.5)
+        pre_pool: whether pre-pool the features or not (True for Aesthetic Attributes, False for Technical Attributes)
+    """
+
+    def __init__(
+        self, in_channels=768, hidden_channels=64, dropout_ratio=0.5, pre_pool=False, **kwargs
+    ):
+        super().__init__()
+        self.dropout_ratio = dropout_ratio
+        self.in_channels = in_channels
+        self.hidden_channels = hidden_channels
+        self.pre_pool = pre_pool
+        if self.dropout_ratio != 0:
+            self.dropout = nn.Dropout(p=self.dropout_ratio)
+        else:
+            self.dropout = None
+        self.fc_hid = nn.Conv3d(self.in_channels, self.hidden_channels, (1, 1, 1))
+        self.fc_last = nn.Conv3d(self.hidden_channels, 1, (1, 1, 1))
+        self.gelu = nn.GELU()
+
+        self.avg_pool = nn.AdaptiveAvgPool3d((1, 1, 1))
+
+    def forward(self, x, rois=None):
+        if self.pre_pool:
+            x = self.avg_pool(x)
+        x = self.dropout(x)
+        qlt_score = self.fc_last(self.dropout(self.gelu(self.fc_hid(x))))
+        return qlt_score
+    
+    
+
+
+
+class VARHead(nn.Module):
+    """MLP Regression Head for Video Action Recognition.
+    Args:
+        in_channels: input channels for MLP
+        hidden_channels: hidden channels for MLP
+        dropout_ratio: the dropout ratio for features before the MLP (default 0.5)
+    """
+
+    def __init__(self, in_channels=768, out_channels=400, dropout_ratio=0.5, **kwargs):
+        super().__init__()
+        self.dropout_ratio = dropout_ratio
+        self.in_channels = in_channels
+        self.out_channels = out_channels
+        if self.dropout_ratio != 0:
+            self.dropout = nn.Dropout(p=self.dropout_ratio)
+        else:
+            self.dropout = None
+        self.fc = nn.Conv3d(self.in_channels, self.out_channels, (1, 1, 1))
+        self.avg_pool = nn.AdaptiveAvgPool3d((1, 1, 1))
+
+    def forward(self, x, rois=None):
+        x = self.dropout(x)
+        x = self.avg_pool(x)
+        out = self.fc(x)
+        return out
+
+
+class IQAHead(nn.Module):
+    """MLP Regression Head for IQA.
+    Args:
+        in_channels: input channels for MLP
+        hidden_channels: hidden channels for MLP
+        dropout_ratio: the dropout ratio for features before the MLP (default 0.5)
+    """
+
+    def __init__(
+        self, in_channels=768, hidden_channels=64, dropout_ratio=0.5, **kwargs
+    ):
+        super().__init__()
+        self.dropout_ratio = dropout_ratio
+        self.in_channels = in_channels
+        self.hidden_channels = hidden_channels
+        if self.dropout_ratio != 0:
+            self.dropout = nn.Dropout(p=self.dropout_ratio)
+        else:
+            self.dropout = None
+        self.fc_hid = nn.Linear(self.in_channels, self.hidden_channels)
+        self.fc_last = nn.Linear(self.hidden_channels, 1)
+        self.gelu = nn.GELU()
+
+    def forward(self, x):
+        x = self.dropout(x)
+        qlt_score = self.fc_last(self.dropout(self.gelu(self.fc_hid(x))))
+        return qlt_score

cover/models/swin_backbone.py

@@ -0,0 +1,1097 @@
+import math
+from functools import lru_cache, reduce
+from operator import mul
+
+import numpy as np
+import torch
+import torch.nn as nn
+import torch.nn.functional as F
+import torch.utils.checkpoint as checkpoint
+from einops import rearrange
+from timm.models.layers import DropPath, trunc_normal_
+
+
+def fragment_infos(D, H, W, fragments=7, device="cuda"):
+    m = torch.arange(fragments).unsqueeze(-1).float()
+    m = (m + m.t() * fragments).reshape(1, 1, 1, fragments, fragments)
+    m = F.interpolate(m.to(device), size=(D, H, W)).permute(0, 2, 3, 4, 1)
+    return m.long()
+
+
+@lru_cache
+def global_position_index(
+    D,
+    H,
+    W,
+    fragments=(1, 7, 7),
+    window_size=(8, 7, 7),
+    shift_size=(0, 0, 0),
+    device="cuda",
+):
+    frags_d = torch.arange(fragments[0])
+    frags_h = torch.arange(fragments[1])
+    frags_w = torch.arange(fragments[2])
+    frags = torch.stack(
+        torch.meshgrid(frags_d, frags_h, frags_w)
+    ).float()  # 3, Fd, Fh, Fw
+    coords = (
+        torch.nn.functional.interpolate(frags[None].to(device), size=(D, H, W))
+        .long()
+        .permute(0, 2, 3, 4, 1)
+    )
+    # print(shift_size)
+    coords = torch.roll(
+        coords, shifts=(-shift_size[0], -shift_size[1], -shift_size[2]), dims=(1, 2, 3)
+    )
+    window_coords = window_partition(coords, window_size)
+    relative_coords = (
+        window_coords[:, None, :] - window_coords[:, :, None]
+    )  # Wd*Wh*Ww, Wd*Wh*Ww, 3
+    return relative_coords  # relative_coords
+
+
+@lru_cache
+def get_adaptive_window_size(
+    base_window_size, input_x_size, base_x_size,
+):
+    tw, hw, ww = base_window_size
+    tx_, hx_, wx_ = input_x_size
+    tx, hx, wx = base_x_size
+    print((tw * tx_) // tx, (hw * hx_) // hx, (ww * wx_) // wx)
+    return (tw * tx_) // tx, (hw * hx_) // hx, (ww * wx_) // wx
+
+
+class Mlp(nn.Module):
+    """Multilayer perceptron."""
+
+    def __init__(
+        self,
+        in_features,
+        hidden_features=None,
+        out_features=None,
+        act_layer=nn.GELU,
+        drop=0.0,
+    ):
+        super().__init__()
+        out_features = out_features or in_features
+        hidden_features = hidden_features or in_features
+        self.fc1 = nn.Linear(in_features, hidden_features)
+        self.act = act_layer()
+        self.fc2 = nn.Linear(hidden_features, out_features)
+        self.drop = nn.Dropout(drop)
+
+    def forward(self, x):
+        x = self.fc1(x)
+        x = self.act(x)
+        x = self.drop(x)
+        x = self.fc2(x)
+        x = self.drop(x)
+        return x
+
+
+def window_partition(x, window_size):
+    """
+    Args:
+        x: (B, D, H, W, C)
+        window_size (tuple[int]): window size
+
+    Returns:
+        windows: (B*num_windows, window_size*window_size, C)
+    """
+    B, D, H, W, C = x.shape
+    x = x.view(
+        B,
+        D // window_size[0],
+        window_size[0],
+        H // window_size[1],
+        window_size[1],
+        W // window_size[2],
+        window_size[2],
+        C,
+    )
+    windows = (
+        x.permute(0, 1, 3, 5, 2, 4, 6, 7)
+        .contiguous()
+        .view(-1, reduce(mul, window_size), C)
+    )
+    return windows
+
+
+def window_reverse(windows, window_size, B, D, H, W):
+    """
+    Args:
+        windows: (B*num_windows, window_size, window_size, C)
+        window_size (tuple[int]): Window size
+        H (int): Height of image
+        W (int): Width of image
+
+    Returns:
+        x: (B, D, H, W, C)
+    """
+    x = windows.view(
+        B,
+        D // window_size[0],
+        H // window_size[1],
+        W // window_size[2],
+        window_size[0],
+        window_size[1],
+        window_size[2],
+        -1,
+    )
+    x = x.permute(0, 1, 4, 2, 5, 3, 6, 7).contiguous().view(B, D, H, W, -1)
+    return x
+
+
+def get_window_size(x_size, window_size, shift_size=None):
+    use_window_size = list(window_size)
+    if shift_size is not None:
+        use_shift_size = list(shift_size)
+    for i in range(len(x_size)):
+        if x_size[i] <= window_size[i]:
+            use_window_size[i] = x_size[i]
+            if shift_size is not None:
+                use_shift_size[i] = 0
+
+    if shift_size is None:
+        return tuple(use_window_size)
+    else:
+        return tuple(use_window_size), tuple(use_shift_size)
+
+
+class WindowAttention3D(nn.Module):
+    """Window based multi-head self attention (W-MSA) module with relative position bias.
+    It supports both of shifted and non-shifted window.
+    Args:
+        dim (int): Number of input channels.
+        window_size (tuple[int]): The temporal length, height and width of the window.
+        num_heads (int): Number of attention heads.
+        qkv_bias (bool, optional):  If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set
+        attn_drop (float, optional): Dropout ratio of attention weight. Default: 0.0
+        proj_drop (float, optional): Dropout ratio of output. Default: 0.0
+    """
+
+    def __init__(
+        self,
+        dim,
+        window_size,
+        num_heads,
+        qkv_bias=False,
+        qk_scale=None,
+        attn_drop=0.0,
+        proj_drop=0.0,
+        frag_bias=False,
+    ):
+
+        super().__init__()
+        self.dim = dim
+        self.window_size = window_size  # Wd, Wh, Ww
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+        self.scale = qk_scale or head_dim ** -0.5
+
+        # define a parameter table of relative position bias
+        self.relative_position_bias_table = nn.Parameter(
+            torch.zeros(
+                (2 * window_size[0] - 1)
+                * (2 * window_size[1] - 1)
+                * (2 * window_size[2] - 1),
+                num_heads,
+            )
+        )  # 2*Wd-1 * 2*Wh-1 * 2*Ww-1, nH
+        if frag_bias:
+            self.fragment_position_bias_table = nn.Parameter(
+                torch.zeros(
+                    (2 * window_size[0] - 1)
+                    * (2 * window_size[1] - 1)
+                    * (2 * window_size[2] - 1),
+                    num_heads,
+                )
+            )
+
+        # get pair-wise relative position index for each token inside the window
+        coords_d = torch.arange(self.window_size[0])
+        coords_h = torch.arange(self.window_size[1])
+        coords_w = torch.arange(self.window_size[2])
+        coords = torch.stack(
+            torch.meshgrid(coords_d, coords_h, coords_w)
+        )  # 3, Wd, Wh, Ww
+        coords_flatten = torch.flatten(coords, 1)  # 3, Wd*Wh*Ww
+        relative_coords = (
+            coords_flatten[:, :, None] - coords_flatten[:, None, :]
+        )  # 3, Wd*Wh*Ww, Wd*Wh*Ww
+        relative_coords = relative_coords.permute(
+            1, 2, 0
+        ).contiguous()  # Wd*Wh*Ww, Wd*Wh*Ww, 3
+        relative_coords[:, :, 0] += self.window_size[0] - 1  # shift to start from 0
+        relative_coords[:, :, 1] += self.window_size[1] - 1
+        relative_coords[:, :, 2] += self.window_size[2] - 1
+
+        relative_coords[:, :, 0] *= (2 * self.window_size[1] - 1) * (
+            2 * self.window_size[2] - 1
+        )
+        relative_coords[:, :, 1] *= 2 * self.window_size[2] - 1
+        relative_position_index = relative_coords.sum(-1)  # Wd*Wh*Ww, Wd*Wh*Ww
+        self.register_buffer("relative_position_index", relative_position_index)
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(dim, dim)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+        trunc_normal_(self.relative_position_bias_table, std=0.02)
+        self.softmax = nn.Softmax(dim=-1)
+
+    def forward(self, x, mask=None, fmask=None, resized_window_size=None):
+        """Forward function.
+        Args:
+            x: input features with shape of (num_windows*B, N, C)
+            mask: (0/-inf) mask with shape of (num_windows, N, N) or None
+        """
+        # print(x.shape)
+        B_, N, C = x.shape
+        qkv = (
+            self.qkv(x)
+            .reshape(B_, N, 3, self.num_heads, C // self.num_heads)
+            .permute(2, 0, 3, 1, 4)
+        )
+        q, k, v = qkv[0], qkv[1], qkv[2]  # B_, nH, N, C
+
+        q = q * self.scale
+        attn = q @ k.transpose(-2, -1)
+
+        if resized_window_size is None:
+            rpi = self.relative_position_index[:N, :N]
+        else:
+            relative_position_index = self.relative_position_index.reshape(
+                *self.window_size, *self.window_size
+            )
+            d, h, w = resized_window_size
+
+            rpi = relative_position_index[:d, :h, :w, :d, :h, :w]
+        relative_position_bias = self.relative_position_bias_table[
+            rpi.reshape(-1)
+        ].reshape(
+            N, N, -1
+        )  # Wd*Wh*Ww,Wd*Wh*Ww,nH
+        relative_position_bias = relative_position_bias.permute(
+            2, 0, 1
+        ).contiguous()  # nH, Wd*Wh*Ww, Wd*Wh*Ww
+        if hasattr(self, "fragment_position_bias_table"):
+            fragment_position_bias = self.fragment_position_bias_table[
+                rpi.reshape(-1)
+            ].reshape(
+                N, N, -1
+            )  # Wd*Wh*Ww,Wd*Wh*Ww,nH
+            fragment_position_bias = fragment_position_bias.permute(
+                2, 0, 1
+            ).contiguous()  # nH, Wd*Wh*Ww, Wd*Wh*Ww
+
+        ### Mask Position Bias
+        if fmask is not None:
+            # fgate = torch.where(fmask - fmask.transpose(-1, -2) == 0, 1, 0).float()
+            fgate = fmask.abs().sum(-1)
+            nW = fmask.shape[0]
+            relative_position_bias = relative_position_bias.unsqueeze(0)
+            fgate = fgate.unsqueeze(1)
+            # print(fgate.shape, relative_position_bias.shape)
+            if hasattr(self, "fragment_position_bias_table"):
+                relative_position_bias = (
+                    relative_position_bias * fgate
+                    + fragment_position_bias * (1 - fgate)
+                )
+
+            attn = attn.view(
+                B_ // nW, nW, self.num_heads, N, N
+            ) + relative_position_bias.unsqueeze(0)
+            attn = attn.view(-1, self.num_heads, N, N)
+        else:
+            attn = attn + relative_position_bias.unsqueeze(0)  # B_, nH, N, N
+
+        if mask is not None:
+            nW = mask.shape[0]
+            attn = attn.view(B_ // nW, nW, self.num_heads, N, N) + mask.unsqueeze(
+                1
+            ).unsqueeze(0)
+            attn = attn.view(-1, self.num_heads, N, N)
+            attn = self.softmax(attn)
+        else:
+            attn = self.softmax(attn)
+        attn = self.attn_drop(attn)
+
+        x = (attn @ v).transpose(1, 2).reshape(B_, N, C)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+
+        return x
+
+
+class SwinTransformerBlock3D(nn.Module):
+    """Swin Transformer Block.
+
+    Args:
+        dim (int): Number of input channels.
+        num_heads (int): Number of attention heads.
+        window_size (tuple[int]): Window size.
+        shift_size (tuple[int]): Shift size for SW-MSA.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float, optional): Stochastic depth rate. Default: 0.0
+        act_layer (nn.Module, optional): Activation layer. Default: nn.GELU
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+
+    def __init__(
+        self,
+        dim,
+        num_heads,
+        window_size=(2, 7, 7),
+        shift_size=(0, 0, 0),
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        qk_scale=None,
+        drop=0.0,
+        attn_drop=0.0,
+        drop_path=0.0,
+        act_layer=nn.GELU,
+        norm_layer=nn.LayerNorm,
+        use_checkpoint=False,
+        jump_attention=False,
+        frag_bias=False,
+    ):
+        super().__init__()
+        self.dim = dim
+        self.num_heads = num_heads
+        self.window_size = window_size
+        self.shift_size = shift_size
+        self.mlp_ratio = mlp_ratio
+        self.use_checkpoint = use_checkpoint
+        self.jump_attention = jump_attention
+        self.frag_bias = frag_bias
+
+        assert (
+            0 <= self.shift_size[0] < self.window_size[0]
+        ), "shift_size must in 0-window_size"
+        assert (
+            0 <= self.shift_size[1] < self.window_size[1]
+        ), "shift_size must in 0-window_size"
+        assert (
+            0 <= self.shift_size[2] < self.window_size[2]
+        ), "shift_size must in 0-window_size"
+
+        self.norm1 = norm_layer(dim)
+        self.attn = WindowAttention3D(
+            dim,
+            window_size=self.window_size,
+            num_heads=num_heads,
+            qkv_bias=qkv_bias,
+            qk_scale=qk_scale,
+            attn_drop=attn_drop,
+            proj_drop=drop,
+            frag_bias=frag_bias,
+        )
+
+        self.drop_path = DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
+        self.norm2 = norm_layer(dim)
+        mlp_hidden_dim = int(dim * mlp_ratio)
+        self.mlp = Mlp(
+            in_features=dim,
+            hidden_features=mlp_hidden_dim,
+            act_layer=act_layer,
+            drop=drop,
+        )
+
+    def forward_part1(self, x, mask_matrix, resized_window_size=None):
+        B, D, H, W, C = x.shape
+        window_size, shift_size = get_window_size(
+            (D, H, W),
+            self.window_size if resized_window_size is None else resized_window_size,
+            self.shift_size,
+        )
+
+        x = self.norm1(x)
+        # pad feature maps to multiples of window size
+        pad_l = pad_t = pad_d0 = 0
+        pad_d1 = (window_size[0] - D % window_size[0]) % window_size[0]
+        pad_b = (window_size[1] - H % window_size[1]) % window_size[1]
+        pad_r = (window_size[2] - W % window_size[2]) % window_size[2]
+
+        x = F.pad(x, (0, 0, pad_l, pad_r, pad_t, pad_b, pad_d0, pad_d1))
+        _, Dp, Hp, Wp, _ = x.shape
+        if False:  # not hasattr(self, 'finfo_windows'):
+            finfo = fragment_infos(Dp, Hp, Wp)
+
+        # cyclic shift
+        if any(i > 0 for i in shift_size):
+            shifted_x = torch.roll(
+                x,
+                shifts=(-shift_size[0], -shift_size[1], -shift_size[2]),
+                dims=(1, 2, 3),
+            )
+            if False:  # not hasattr(self, 'finfo_windows'):
+                shifted_finfo = torch.roll(
+                    finfo,
+                    shifts=(-shift_size[0], -shift_size[1], -shift_size[2]),
+                    dims=(1, 2, 3),
+                )
+            attn_mask = mask_matrix
+        else:
+            shifted_x = x
+            if False:  # not hasattr(self, 'finfo_windows'):
+                shifted_finfo = finfo
+            attn_mask = None
+        # partition windows
+        x_windows = window_partition(shifted_x, window_size)  # B*nW, Wd*Wh*Ww, C
+        if False:  # not hasattr(self, 'finfo_windows'):
+            self.finfo_windows = window_partition(shifted_finfo, window_size)
+        # W-MSA/SW-MSA
+        # print(shift_size)
+        gpi = global_position_index(
+            Dp,
+            Hp,
+            Wp,
+            fragments=(1,) + window_size[1:],
+            window_size=window_size,
+            shift_size=shift_size,
+            device=x.device,
+        )
+        attn_windows = self.attn(
+            x_windows,
+            mask=attn_mask,
+            fmask=gpi,
+            resized_window_size=window_size
+            if resized_window_size is not None
+            else None,
+        )  # self.finfo_windows)  # B*nW, Wd*Wh*Ww, C
+        # merge windows
+        attn_windows = attn_windows.view(-1, *(window_size + (C,)))
+        shifted_x = window_reverse(
+            attn_windows, window_size, B, Dp, Hp, Wp
+        )  # B D' H' W' C
+        # reverse cyclic shift
+        if any(i > 0 for i in shift_size):
+            x = torch.roll(
+                shifted_x,
+                shifts=(shift_size[0], shift_size[1], shift_size[2]),
+                dims=(1, 2, 3),
+            )
+        else:
+            x = shifted_x
+
+        if pad_d1 > 0 or pad_r > 0 or pad_b > 0:
+            x = x[:, :D, :H, :W, :].contiguous()
+        return x
+
+    def forward_part2(self, x):
+        return self.drop_path(self.mlp(self.norm2(x)))
+
+    def forward(self, x, mask_matrix, resized_window_size=None):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, D, H, W, C).
+            mask_matrix: Attention mask for cyclic shift.
+        """
+
+        shortcut = x
+        if not self.jump_attention:
+            if self.use_checkpoint:
+                x = checkpoint.checkpoint(
+                    self.forward_part1, x, mask_matrix, resized_window_size
+                )
+            else:
+                x = self.forward_part1(x, mask_matrix, resized_window_size)
+            x = shortcut + self.drop_path(x)
+
+        if self.use_checkpoint:
+            x = x + checkpoint.checkpoint(self.forward_part2, x)
+        else:
+            x = x + self.forward_part2(x)
+
+        return x
+
+
+class PatchMerging(nn.Module):
+    """Patch Merging Layer
+
+    Args:
+        dim (int): Number of input channels.
+        norm_layer (nn.Module, optional): Normalization layer.  Default: nn.LayerNorm
+    """
+
+    def __init__(self, dim, norm_layer=nn.LayerNorm):
+        super().__init__()
+        self.dim = dim
+        self.reduction = nn.Linear(4 * dim, 2 * dim, bias=False)
+        self.norm = norm_layer(4 * dim)
+
+    def forward(self, x):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, D, H, W, C).
+        """
+        B, D, H, W, C = x.shape
+
+        # padding
+        pad_input = (H % 2 == 1) or (W % 2 == 1)
+        if pad_input:
+            x = F.pad(x, (0, 0, 0, W % 2, 0, H % 2))
+
+        x0 = x[:, :, 0::2, 0::2, :]  # B D H/2 W/2 C
+        x1 = x[:, :, 1::2, 0::2, :]  # B D H/2 W/2 C
+        x2 = x[:, :, 0::2, 1::2, :]  # B D H/2 W/2 C
+        x3 = x[:, :, 1::2, 1::2, :]  # B D H/2 W/2 C
+        x = torch.cat([x0, x1, x2, x3], -1)  # B D H/2 W/2 4*C
+
+        x = self.norm(x)
+        x = self.reduction(x)
+
+        return x
+
+
+# cache each stage results
+@lru_cache()
+def compute_mask(D, H, W, window_size, shift_size, device):
+    img_mask = torch.zeros((1, D, H, W, 1), device=device)  # 1 Dp Hp Wp 1
+    cnt = 0
+    for d in (
+        slice(-window_size[0]),
+        slice(-window_size[0], -shift_size[0]),
+        slice(-shift_size[0], None),
+    ):
+        for h in (
+            slice(-window_size[1]),
+            slice(-window_size[1], -shift_size[1]),
+            slice(-shift_size[1], None),
+        ):
+            for w in (
+                slice(-window_size[2]),
+                slice(-window_size[2], -shift_size[2]),
+                slice(-shift_size[2], None),
+            ):
+                img_mask[:, d, h, w, :] = cnt
+                cnt += 1
+    mask_windows = window_partition(img_mask, window_size)  # nW, ws[0]*ws[1]*ws[2], 1
+    mask_windows = mask_windows.squeeze(-1)  # nW, ws[0]*ws[1]*ws[2]
+    attn_mask = mask_windows.unsqueeze(1) - mask_windows.unsqueeze(2)
+    attn_mask = attn_mask.masked_fill(attn_mask != 0, float(-100.0)).masked_fill(
+        attn_mask == 0, float(0.0)
+    )
+    return attn_mask
+
+
+class BasicLayer(nn.Module):
+    """A basic Swin Transformer layer for one stage.
+
+    Args:
+        dim (int): Number of feature channels
+        depth (int): Depths of this stage.
+        num_heads (int): Number of attention head.
+        window_size (tuple[int]): Local window size. Default: (1,7,7).
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool, optional): If True, add a learnable bias to query, key, value. Default: True
+        qk_scale (float | None, optional): Override default qk scale of head_dim ** -0.5 if set.
+        drop (float, optional): Dropout rate. Default: 0.0
+        attn_drop (float, optional): Attention dropout rate. Default: 0.0
+        drop_path (float | tuple[float], optional): Stochastic depth rate. Default: 0.0
+        norm_layer (nn.Module, optional): Normalization layer. Default: nn.LayerNorm
+        downsample (nn.Module | None, optional): Downsample layer at the end of the layer. Default: None
+    """
+
+    def __init__(
+        self,
+        dim,
+        depth,
+        num_heads,
+        window_size=(1, 7, 7),
+        mlp_ratio=4.0,
+        qkv_bias=False,
+        qk_scale=None,
+        drop=0.0,
+        attn_drop=0.0,
+        drop_path=0.0,
+        norm_layer=nn.LayerNorm,
+        downsample=None,
+        use_checkpoint=False,
+        jump_attention=False,
+        frag_bias=False,
+    ):
+        super().__init__()
+        self.window_size = window_size
+        self.shift_size = tuple(i // 2 for i in window_size)
+        self.depth = depth
+        self.use_checkpoint = use_checkpoint
+        # print(window_size)
+        # build blocks
+        self.blocks = nn.ModuleList(
+            [
+                SwinTransformerBlock3D(
+                    dim=dim,
+                    num_heads=num_heads,
+                    window_size=window_size,
+                    shift_size=(0, 0, 0) if (i % 2 == 0) else self.shift_size,
+                    mlp_ratio=mlp_ratio,
+                    qkv_bias=qkv_bias,
+                    qk_scale=qk_scale,
+                    drop=drop,
+                    attn_drop=attn_drop,
+                    drop_path=drop_path[i]
+                    if isinstance(drop_path, list)
+                    else drop_path,
+                    norm_layer=norm_layer,
+                    use_checkpoint=use_checkpoint,
+                    jump_attention=jump_attention,
+                    frag_bias=frag_bias,
+                )
+                for i in range(depth)
+            ]
+        )
+
+        self.downsample = downsample
+        if self.downsample is not None:
+            self.downsample = downsample(dim=dim, norm_layer=norm_layer)
+
+    def forward(self, x, resized_window_size=None):
+        """Forward function.
+
+        Args:
+            x: Input feature, tensor size (B, C, D, H, W).
+        """
+        # calculate attention mask for SW-MSA
+        B, C, D, H, W = x.shape
+
+        window_size, shift_size = get_window_size(
+            (D, H, W),
+            self.window_size if resized_window_size is None else resized_window_size,
+            self.shift_size,
+        )
+        # print(window_size)
+        x = rearrange(x, "b c d h w -> b d h w c")
+        Dp = int(np.ceil(D / window_size[0])) * window_size[0]
+        Hp = int(np.ceil(H / window_size[1])) * window_size[1]
+        Wp = int(np.ceil(W / window_size[2])) * window_size[2]
+        attn_mask = compute_mask(Dp, Hp, Wp, window_size, shift_size, x.device)
+        for blk in self.blocks:
+            x = blk(x, attn_mask, resized_window_size=resized_window_size)
+        x = x.view(B, D, H, W, -1)
+
+        if self.downsample is not None:
+            x = self.downsample(x)
+        x = rearrange(x, "b d h w c -> b c d h w")
+        return x
+
+
+class PatchEmbed3D(nn.Module):
+    """Video to Patch Embedding.
+
+    Args:
+        patch_size (int): Patch token size. Default: (2,4,4).
+        in_chans (int): Number of input video channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        norm_layer (nn.Module, optional): Normalization layer. Default: None
+    """
+
+    def __init__(self, patch_size=(2, 4, 4), in_chans=3, embed_dim=96, norm_layer=None):
+        super().__init__()
+        self.patch_size = patch_size
+
+        self.in_chans = in_chans
+        self.embed_dim = embed_dim
+
+        self.proj = nn.Conv3d(
+            in_chans, embed_dim, kernel_size=patch_size, stride=patch_size
+        )
+        if norm_layer is not None:
+            self.norm = norm_layer(embed_dim)
+        else:
+            self.norm = None
+
+    def forward(self, x):
+        """Forward function."""
+        # padding
+        _, _, D, H, W = x.size()
+        if W % self.patch_size[2] != 0:
+            x = F.pad(x, (0, self.patch_size[2] - W % self.patch_size[2]))
+        if H % self.patch_size[1] != 0:
+            x = F.pad(x, (0, 0, 0, self.patch_size[1] - H % self.patch_size[1]))
+        if D % self.patch_size[0] != 0:
+            x = F.pad(x, (0, 0, 0, 0, 0, self.patch_size[0] - D % self.patch_size[0]))
+
+        x = self.proj(x)  # B C D Wh Ww
+        if self.norm is not None:
+            D, Wh, Ww = x.size(2), x.size(3), x.size(4)
+            x = x.flatten(2).transpose(1, 2)
+            x = self.norm(x)
+            x = x.transpose(1, 2).view(-1, self.embed_dim, D, Wh, Ww)
+
+        return x
+
+
+class SwinTransformer3D(nn.Module):
+    """Swin Transformer backbone.
+        A PyTorch impl of : `Swin Transformer: Hierarchical Vision Transformer using Shifted Windows`  -
+          https://arxiv.org/pdf/2103.14030
+
+    Args:
+        patch_size (int | tuple(int)): Patch size. Default: (4,4,4).
+        in_chans (int): Number of input image channels. Default: 3.
+        embed_dim (int): Number of linear projection output channels. Default: 96.
+        depths (tuple[int]): Depths of each Swin Transformer stage.
+        num_heads (tuple[int]): Number of attention head of each stage.
+        window_size (int): Window size. Default: 7.
+        mlp_ratio (float): Ratio of mlp hidden dim to embedding dim. Default: 4.
+        qkv_bias (bool): If True, add a learnable bias to query, key, value. Default: Truee
+        qk_scale (float): Override default qk scale of head_dim ** -0.5 if set.
+        drop_rate (float): Dropout rate.
+        attn_drop_rate (float): Attention dropout rate. Default: 0.
+        drop_path_rate (float): Stochastic depth rate. Default: 0.2.
+        norm_layer: Normalization layer. Default: nn.LayerNorm.
+        patch_norm (bool): If True, add normalization after patch embedding. Default: False.
+        frozen_stages (int): Stages to be frozen (stop grad and set eval mode).
+            -1 means not freezing any parameters.
+    """
+
+    def __init__(
+        self,
+        pretrained=None,
+        pretrained2d=False,
+        patch_size=(2, 4, 4),
+        in_chans=3,
+        embed_dim=96,
+        depths=[2, 2, 6, 2],
+        num_heads=[3, 6, 12, 24],
+        window_size=(8, 7, 7),
+        mlp_ratio=4.0,
+        qkv_bias=True,
+        qk_scale=None,
+        drop_rate=0.0,
+        attn_drop_rate=0.0,
+        drop_path_rate=0.1,
+        norm_layer=nn.LayerNorm,
+        patch_norm=True,
+        frozen_stages=-1,
+        use_checkpoint=True,
+        jump_attention=[False, False, False, False],
+        frag_biases=[True, True, True, False],
+        base_x_size=(32, 224, 224),
+    ):
+        super().__init__()
+
+        self.pretrained = pretrained
+        self.pretrained2d = pretrained2d
+        self.num_layers = len(depths)
+        self.embed_dim = embed_dim
+        self.patch_norm = patch_norm
+        self.frozen_stages = frozen_stages
+        self.window_size = window_size
+        self.patch_size = patch_size
+        self.base_x_size = base_x_size
+
+        # split image into non-overlapping patches
+        self.patch_embed = PatchEmbed3D(
+            patch_size=patch_size,
+            in_chans=in_chans,
+            embed_dim=embed_dim,
+            norm_layer=norm_layer if self.patch_norm else None,
+        )
+
+        self.pos_drop = nn.Dropout(p=drop_rate)
+
+        # stochastic depth
+        dpr = [
+            x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))
+        ]  # stochastic depth decay rule
+
+        # build layers
+        self.layers = nn.ModuleList()
+        for i_layer in range(self.num_layers):
+            layer = BasicLayer(
+                dim=int(embed_dim * 2 ** i_layer),
+                depth=depths[i_layer],
+                num_heads=num_heads[i_layer],
+                window_size=window_size[i_layer]
+                if isinstance(window_size, list)
+                else window_size,
+                mlp_ratio=mlp_ratio,
+                qkv_bias=qkv_bias,
+                qk_scale=qk_scale,
+                drop=drop_rate,
+                attn_drop=attn_drop_rate,
+                drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])],
+                norm_layer=norm_layer,
+                downsample=PatchMerging if i_layer < self.num_layers - 1 else None,
+                use_checkpoint=use_checkpoint,
+                jump_attention=jump_attention[i_layer],
+                frag_bias=frag_biases[i_layer],
+            )
+            self.layers.append(layer)
+
+        self.num_features = int(embed_dim * 2 ** (self.num_layers - 1))
+
+        # add a norm layer for each output
+        self.norm = norm_layer(self.num_features)
+
+        self._freeze_stages()
+
+        self.init_weights()
+
+    def _freeze_stages(self):
+        if self.frozen_stages >= 0:
+            self.patch_embed.eval()
+            for param in self.patch_embed.parameters():
+                param.requires_grad = False
+
+        if self.frozen_stages >= 1:
+            self.pos_drop.eval()
+            for i in range(0, self.frozen_stages):
+                m = self.layers[i]
+                m.eval()
+                for param in m.parameters():
+                    param.requires_grad = False
+
+    def inflate_weights(self):
+        """Inflate the swin2d parameters to swin3d.
+
+        The differences between swin3d and swin2d mainly lie in an extra
+        axis. To utilize the pretrained parameters in 2d model,
+        the weight of swin2d models should be inflated to fit in the shapes of
+        the 3d counterpart.
+
+        Args:
+            logger (logging.Logger): The logger used to print
+                debugging infomation.
+        """
+        checkpoint = torch.load(self.pretrained, map_location="cpu")
+        state_dict = checkpoint["model"]
+
+        # delete relative_position_index since we always re-init it
+        relative_position_index_keys = [
+            k for k in state_dict.keys() if "relative_position_index" in k
+        ]
+        for k in relative_position_index_keys:
+            del state_dict[k]
+
+        # delete attn_mask since we always re-init it
+        attn_mask_keys = [k for k in state_dict.keys() if "attn_mask" in k]
+        for k in attn_mask_keys:
+            del state_dict[k]
+
+        state_dict["patch_embed.proj.weight"] = (
+            state_dict["patch_embed.proj.weight"]
+            .unsqueeze(2)
+            .repeat(1, 1, self.patch_size[0], 1, 1)
+            / self.patch_size[0]
+        )
+
+        # bicubic interpolate relative_position_bias_table if not match
+        relative_position_bias_table_keys = [
+            k for k in state_dict.keys() if "relative_position_bias_table" in k
+        ]
+        for k in relative_position_bias_table_keys:
+            relative_position_bias_table_pretrained = state_dict[k]
+            relative_position_bias_table_current = self.state_dict()[k]
+            L1, nH1 = relative_position_bias_table_pretrained.size()
+            L2, nH2 = relative_position_bias_table_current.size()
+            L2 = (2 * self.window_size[1] - 1) * (2 * self.window_size[2] - 1)
+            wd = self.window_size[0]
+            if nH1 != nH2:
+                print(f"Error in loading {k}, passing")
+            else:
+                if L1 != L2:
+                    S1 = int(L1 ** 0.5)
+                    relative_position_bias_table_pretrained_resized = torch.nn.functional.interpolate(
+                        relative_position_bias_table_pretrained.permute(1, 0).view(
+                            1, nH1, S1, S1
+                        ),
+                        size=(
+                            2 * self.window_size[1] - 1,
+                            2 * self.window_size[2] - 1,
+                        ),
+                        mode="bicubic",
+                    )
+                    relative_position_bias_table_pretrained = relative_position_bias_table_pretrained_resized.view(
+                        nH2, L2
+                    ).permute(
+                        1, 0
+                    )
+            state_dict[k] = relative_position_bias_table_pretrained.repeat(
+                2 * wd - 1, 1
+            )
+
+        msg = self.load_state_dict(state_dict, strict=False)
+        print(msg)
+        print(f"=> loaded successfully '{self.pretrained}'")
+        del checkpoint
+        torch.cuda.empty_cache()
+
+    def load_swin(self, load_path, strict=False):
+        print("loading swin lah")
+        from collections import OrderedDict
+
+        model_state_dict = self.state_dict()
+        state_dict = torch.load(load_path)["state_dict"]
+
+        clean_dict = OrderedDict()
+        for key, value in state_dict.items():
+            if "backbone" in key:
+                clean_key = key[9:]
+                clean_dict[clean_key] = value
+                if "relative_position_bias_table" in clean_key:
+                    forked_key = clean_key.replace(
+                        "relative_position_bias_table", "fragment_position_bias_table"
+                    )
+                    if forked_key in clean_dict:
+                        print("load_swin_error?")
+                    else:
+                        clean_dict[forked_key] = value
+
+        # bicubic interpolate relative_position_bias_table if not match
+        relative_position_bias_table_keys = [
+            k for k in clean_dict.keys() if "relative_position_bias_table" in k
+        ]
+        for k in relative_position_bias_table_keys:
+            print(k)
+            relative_position_bias_table_pretrained = clean_dict[k]
+            relative_position_bias_table_current = model_state_dict[k]
+            L1, nH1 = relative_position_bias_table_pretrained.size()
+            L2, nH2 = relative_position_bias_table_current.size()
+            if isinstance(self.window_size, list):
+                i_layer = int(k.split(".")[1])
+                L2 = (2 * self.window_size[i_layer][1] - 1) * (
+                    2 * self.window_size[i_layer][2] - 1
+                )
+                wd = self.window_size[i_layer][0]
+            else:
+                L2 = (2 * self.window_size[1] - 1) * (2 * self.window_size[2] - 1)
+                wd = self.window_size[0]
+            if nH1 != nH2:
+                print(f"Error in loading {k}, passing")
+            else:
+                if L1 != L2:
+                    S1 = int((L1 / 15) ** 0.5)
+                    print(
+                        relative_position_bias_table_pretrained.shape, 15, nH1, S1, S1
+                    )
+                    relative_position_bias_table_pretrained_resized = torch.nn.functional.interpolate(
+                        relative_position_bias_table_pretrained.permute(1, 0)
+                        .view(nH1, 15, S1, S1)
+                        .transpose(0, 1),
+                        size=(
+                            2 * self.window_size[i_layer][1] - 1,
+                            2 * self.window_size[i_layer][2] - 1,
+                        ),
+                        mode="bicubic",
+                    )
+                    relative_position_bias_table_pretrained = relative_position_bias_table_pretrained_resized.transpose(
+                        0, 1
+                    ).view(
+                        nH2, 15, L2
+                    )
+            clean_dict[k] = relative_position_bias_table_pretrained  # .repeat(2*wd-1,1)
+
+        ## Clean Mismatched Keys
+        for key, value in model_state_dict.items():
+            if key in clean_dict:
+                if value.shape != clean_dict[key].shape:
+                    print(key)
+                    clean_dict.pop(key)
+
+        self.load_state_dict(clean_dict, strict=strict)
+
+    def init_weights(self, pretrained=None):
+        print(self.pretrained, self.pretrained2d)
+        """Initialize the weights in backbone.
+
+        Args:
+            pretrained (str, optional): Path to pre-trained weights.
+                Defaults to None.
+        """
+
+        def _init_weights(m):
+            if isinstance(m, nn.Linear):
+                trunc_normal_(m.weight, std=0.02)
+                if isinstance(m, nn.Linear) and m.bias is not None:
+                    nn.init.constant_(m.bias, 0)
+            elif isinstance(m, nn.LayerNorm):
+                nn.init.constant_(m.bias, 0)
+                nn.init.constant_(m.weight, 1.0)
+
+        if pretrained:
+            self.pretrained = pretrained
+        if isinstance(self.pretrained, str):
+            self.apply(_init_weights)
+            # logger = get_root_logger()
+            # logger.info(f"load model from: {self.pretrained}")
+
+            if self.pretrained2d:
+                # Inflate 2D model into 3D model.
+                self.inflate_weights()
+            else:
+                # Directly load 3D model.
+                self.load_swin(self.pretrained, strict=False)  # , logger=logger)
+        elif self.pretrained is None:
+            self.apply(_init_weights)
+        else:
+            raise TypeError("pretrained must be a str or None")
+
+    def forward(self, x, multi=False, layer=-1, adaptive_window_size=False):
+
+        """Forward function."""
+        if adaptive_window_size:
+            resized_window_size = get_adaptive_window_size(
+                self.window_size, x.shape[2:], self.base_x_size
+            )
+        else:
+            resized_window_size = None
+
+        x = self.patch_embed(x)
+
+        x = self.pos_drop(x)
+        feats = [x]
+
+        for l, mlayer in enumerate(self.layers):
+            x = mlayer(x.contiguous(), resized_window_size)
+            feats += [x]
+
+        x = rearrange(x, "n c d h w -> n d h w c")
+        x = self.norm(x)
+        x = rearrange(x, "n d h w c -> n c d h w")
+
+        if multi:
+            shape = x.shape[2:]
+            return torch.cat(
+                [F.interpolate(xi, size=shape, mode="trilinear") for xi in feats[:-1]],
+                1,
+            )
+        elif layer > -1:
+            print("something", len(feats))
+            return feats[layer]
+        else:
+            return x
+
+    def train(self, mode=True):
+        """Convert the model into training mode while keep layers freezed."""
+        super(SwinTransformer3D, self).train(mode)
+        self._freeze_stages()
+
+
+def swin_3d_tiny(**kwargs):
+    ## Original Swin-3D Tiny with reduced windows
+    return SwinTransformer3D(depths=[2, 2, 6, 2], frag_biases=[0, 0, 0, 0], **kwargs)
+
+
+def swin_3d_small(**kwargs):
+    # Original Swin-3D Small with reduced windows
+    return SwinTransformer3D(depths=[2, 2, 18, 2], frag_biases=[0, 0, 0, 0], **kwargs)
+
+
+class SwinTransformer2D(nn.Sequential):
+    def __init__(self):
+        ## Only backbone for Swin Transformer 2D
+        from timm.models import swin_tiny_patch4_window7_224
+
+        super().__init__(*list(swin_tiny_patch4_window7_224().children())[:-2])

cover/models/xclip_backbone.py

@@ -0,0 +1,902 @@
+import copy
+import math
+from collections import OrderedDict
+from typing import Tuple, Union
+
+import clip
+import numpy as np
+import torch
+import torch.nn.functional as F
+from einops import rearrange
+from timm.models.layers import trunc_normal_
+from torch import nn
+from torch.utils.checkpoint import checkpoint_sequential
+
+
+def drop_path(x, drop_prob: float = 0.0, training: bool = False):
+    """Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks).
+    This is the same as the DropConnect impl I created for EfficientNet, etc networks, however,
+    the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper...
+    See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for
+    changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use
+    'survival rate' as the argument.
+    """
+    if drop_prob == 0.0 or not training:
+        return x
+    keep_prob = 1 - drop_prob
+    shape = (x.shape[0],) + (1,) * (
+        x.ndim - 1
+    )  # work with diff dim tensors, not just 2D ConvNets
+    random_tensor = keep_prob + torch.rand(shape, dtype=x.dtype, device=x.device)
+    random_tensor.floor_()  # binarize
+    output = x.div(keep_prob) * random_tensor
+    return output
+
+
+class DropPath(nn.Module):
+    """Drop paths (Stochastic Depth) per sample  (when applied in main path of residual blocks)."""
+
+    def __init__(self, drop_prob=None):
+        super(DropPath, self).__init__()
+        self.drop_prob = drop_prob
+
+    def forward(self, x):
+        return drop_path(x, self.drop_prob, self.training)
+
+
+class LayerNorm(nn.LayerNorm):
+    """Subclass torch's LayerNorm to handle fp16."""
+
+    def forward(self, x: torch.Tensor):
+        # orig_type = x.dtype
+        # ret = super().forward(x.type(torch.float32))
+        # return ret.type(orig_type)
+        return super().forward(x)
+
+
+class QuickGELU(nn.Module):
+    def forward(self, x: torch.Tensor):
+        return x * torch.sigmoid(1.702 * x)
+
+
+class ResidualAttentionBlock(nn.Module):
+    def __init__(
+        self, d_model: int, n_head: int, attn_mask: torch.Tensor = None,
+    ):
+        super().__init__()
+
+        self.attn = nn.MultiheadAttention(d_model, n_head,)
+        self.ln_1 = LayerNorm(d_model)
+
+        self.mlp = nn.Sequential(
+            OrderedDict(
+                [
+                    ("c_fc", nn.Linear(d_model, d_model * 4)),
+                    ("gelu", QuickGELU()),
+                    ("c_proj", nn.Linear(d_model * 4, d_model)),
+                ]
+            )
+        )
+        self.ln_2 = LayerNorm(d_model)
+        self.attn_mask = attn_mask
+
+    def attention(self, x: torch.Tensor):
+        self.attn_mask = (
+            self.attn_mask.to(dtype=x.dtype, device=x.device)
+            if self.attn_mask is not None
+            else None
+        )
+        return self.attn(x, x, x, need_weights=False, attn_mask=self.attn_mask)[0]
+
+    def forward(self, x: torch.Tensor):
+        x = x + self.attention(self.ln_1(x))
+        x = x + self.mlp(self.ln_2(x))
+        return x
+
+
+class Transformer(nn.Module):
+    def __init__(
+        self, width: int, layers: int, heads: int, attn_mask: torch.Tensor = None
+    ):
+        super().__init__()
+        self.width = width
+        self.layers = layers
+        self.resblocks = nn.Sequential(
+            *[ResidualAttentionBlock(width, heads, attn_mask) for _ in range(layers)]
+        )
+
+    def forward(self, x: torch.Tensor):
+        return self.resblocks(x)
+
+
+class VisionTransformer(nn.Module):
+    def __init__(
+        self,
+        input_resolution: int,
+        patch_size: int,
+        width: int,
+        layers: int,
+        heads: int,
+        output_dim: int,
+    ):
+        super().__init__()
+        self.input_resolution = input_resolution
+        self.output_dim = output_dim
+        self.conv1 = nn.Conv2d(
+            in_channels=3,
+            out_channels=width,
+            kernel_size=patch_size,
+            stride=patch_size,
+            bias=False,
+        )
+
+        scale = width ** -0.5
+        self.class_embedding = nn.Parameter(scale * torch.randn(width))
+        self.positional_embedding = nn.Parameter(
+            scale * torch.randn((input_resolution // patch_size) ** 2 + 1, width)
+        )
+        self.ln_pre = LayerNorm(width)
+
+        self.transformer = Transformer(width, layers, heads)
+
+        self.ln_post = LayerNorm(width)
+        self.proj = nn.Parameter(scale * torch.randn(width, output_dim))
+
+    def forward(self, x: torch.Tensor):
+        x = self.conv1(x)  # shape = [*, width, grid, grid]
+        x = x.reshape(x.shape[0], x.shape[1], -1)  # shape = [*, width, grid ** 2]
+        x = x.permute(0, 2, 1)  # shape = [*, grid ** 2, width]
+        x = torch.cat(
+            [
+                self.class_embedding.to(x.dtype)
+                + torch.zeros(
+                    x.shape[0], 1, x.shape[-1], dtype=x.dtype, device=x.device
+                ),
+                x,
+            ],
+            dim=1,
+        )  # shape = [*, grid ** 2 + 1, width]
+        x = x + self.positional_embedding.to(x.dtype)
+        x = self.ln_pre(x)
+
+        x = x.permute(1, 0, 2)  # NLD -> LND
+        x = self.transformer(x)
+        x = x.permute(1, 0, 2)  # LND -> NLD
+
+        x = self.ln_post(x[:, 0, :])
+
+        if self.proj is not None:
+            x = x @ self.proj
+        return x
+
+
+class CLIP(nn.Module):
+    def __init__(
+        self,
+        embed_dim: int,
+        # vision
+        image_resolution: int,
+        vision_layers: Union[Tuple[int, int, int, int], int],
+        vision_width: int,
+        vision_patch_size: int,
+        # text
+        context_length: int,
+        vocab_size: int,
+        transformer_width: int,
+        transformer_heads: int,
+        transformer_layers: int,
+    ):
+        super().__init__()
+
+        self.context_length = context_length
+
+        # vision_heads = vision_width // 64
+        # self.visual = VisionTransformer(
+        #     input_resolution=image_resolution,
+        #     patch_size=vision_patch_size,
+        #     width=vision_width,
+        #     layers=vision_layers,
+        #     heads=vision_heads,
+        #     output_dim=embed_dim
+        # )
+
+        # self.transformer = Transformer(
+        #     width=transformer_width,
+        #     layers=transformer_layers,
+        #     heads=transformer_heads,
+        #     attn_mask=self.build_attention_mask()
+        # )
+
+        # self.vocab_size = vocab_size
+        # self.token_embedding = nn.Embedding(vocab_size, transformer_width)
+        # self.positional_embedding = nn.Parameter(torch.empty(self.context_length, transformer_width))
+        # self.ln_final = LayerNorm(transformer_width)
+
+        # self.text_projection = nn.Parameter(torch.empty(transformer_width, embed_dim))
+        # self.logit_scale = nn.Parameter(torch.ones([]) * np.log(1 / 0.07))
+
+        # self.initialize_parameters()
+
+    def initialize_parameters(self):
+        nn.init.normal_(self.token_embedding.weight, std=0.02)
+        nn.init.normal_(self.positional_embedding, std=0.01)
+
+        proj_std = (self.transformer.width ** -0.5) * (
+            (2 * self.transformer.layers) ** -0.5
+        )
+        attn_std = self.transformer.width ** -0.5
+        fc_std = (2 * self.transformer.width) ** -0.5
+        for block in self.transformer.resblocks:
+            nn.init.normal_(block.attn.in_proj_weight, std=attn_std)
+            nn.init.normal_(block.attn.out_proj.weight, std=proj_std)
+            nn.init.normal_(block.mlp.c_fc.weight, std=fc_std)
+            nn.init.normal_(block.mlp.c_proj.weight, std=proj_std)
+
+        if self.text_projection is not None:
+            nn.init.normal_(self.text_projection, std=self.transformer.width ** -0.5)
+
+    def build_attention_mask(self):
+        # lazily create causal attention mask, with full attention between the vision tokens
+        # pytorch uses additive attention mask; fill with -inf
+        mask = torch.empty(self.context_length, self.context_length)
+        mask.fill_(float("-inf"))
+        mask.triu_(1)  # zero out the lower diagonal
+        return mask
+
+    @property
+    def dtype(self):
+        return self.visual.conv1.weight.dtype
+
+    def encode_image(self, image):
+        return self.visual(image.type(self.dtype))
+
+    def encode_text(self, text):
+        x = self.token_embedding(text).type(self.dtype)  # [batch_size, n_ctx, d_model]
+
+        x = x + self.positional_embedding.type(self.dtype)
+        x = x.permute(1, 0, 2)  # NLD -> LND
+        x = self.transformer(x)
+        x = x.permute(1, 0, 2)  # LND -> NLD
+        x = self.ln_final(x).type(self.dtype)
+
+        # x.shape = [batch_size, n_ctx, transformer.width]
+        # take features from the eot embedding (eot_token is the highest number in each sequence)
+        x = x[torch.arange(x.shape[0]), text.argmax(dim=-1)] @ self.text_projection
+
+        return x
+
+    def forward(self, image, text):
+        image_features = self.encode_image(image)
+        text_features = self.encode_text(text)
+
+        # normalized features
+        image_features = image_features / image_features.norm(dim=1, keepdim=True)
+        text_features = text_features / text_features.norm(dim=1, keepdim=True)
+
+        # cosine similarity as logits
+        logit_scale = self.logit_scale.exp()
+        logits_per_image = logit_scale * image_features @ text_features.t()
+        logits_per_text = logits_per_image.t()
+
+        # shape = [global_batch_size, global_batch_size]
+        return logits_per_image, logits_per_text
+
+
+class CrossFramelAttentionBlock(nn.Module):
+    def __init__(
+        self,
+        d_model: int,
+        n_head: int,
+        attn_mask: torch.Tensor = None,
+        droppath=0.0,
+        T=0,
+    ):
+        super().__init__()
+        self.T = T
+
+        self.message_fc = nn.Linear(d_model, d_model)
+        self.message_ln = LayerNorm(d_model)
+        self.message_attn = nn.MultiheadAttention(d_model, n_head,)
+
+        self.attn = nn.MultiheadAttention(d_model, n_head,)
+        self.ln_1 = LayerNorm(d_model)
+
+        self.drop_path = DropPath(droppath) if droppath > 0.0 else nn.Identity()
+        self.mlp = nn.Sequential(
+            OrderedDict(
+                [
+                    ("c_fc", nn.Linear(d_model, d_model * 4)),
+                    ("gelu", QuickGELU()),
+                    ("c_proj", nn.Linear(d_model * 4, d_model)),
+                ]
+            )
+        )
+        self.ln_2 = LayerNorm(d_model)
+        self.attn_mask = attn_mask
+
+    def attention(self, x: torch.Tensor):
+        self.attn_mask = (
+            self.attn_mask.to(dtype=x.dtype, device=x.device)
+            if self.attn_mask is not None
+            else None
+        )
+        return self.attn(x, x, x, need_weights=False, attn_mask=self.attn_mask)[0]
+
+    def forward(self, x):
+        l, bt, d = x.size()
+        b = bt // self.T
+        x = x.view(l, b, self.T, d)
+
+        msg_token = self.message_fc(x[0, :, :, :])
+        msg_token = msg_token.view(b, self.T, 1, d)
+
+        msg_token = msg_token.permute(1, 2, 0, 3).view(self.T, b, d)
+        msg_token = msg_token + self.drop_path(
+            self.message_attn(
+                self.message_ln(msg_token),
+                self.message_ln(msg_token),
+                self.message_ln(msg_token),
+                need_weights=False,
+            )[0]
+        )
+        msg_token = msg_token.view(self.T, 1, b, d).permute(1, 2, 0, 3)
+
+        x = torch.cat([x, msg_token], dim=0)
+
+        x = x.view(l + 1, -1, d)
+        x = x + self.drop_path(self.attention(self.ln_1(x)))
+        x = x[:l, :, :]
+        x = x + self.drop_path(self.mlp(self.ln_2(x)))
+        return x
+
+
+class Transformer(nn.Module):
+    def __init__(
+        self,
+        width: int,
+        layers: int,
+        heads: int,
+        attn_mask: torch.Tensor = None,
+        droppath=None,
+        use_checkpoint=False,
+        T=8,
+    ):
+        super().__init__()
+        self.use_checkpoint = use_checkpoint
+        if droppath is None:
+            droppath = [0.0 for i in range(layers)]
+        self.width = width
+        self.layers = layers
+
+        self.resblocks = nn.Sequential(
+            *[
+                CrossFramelAttentionBlock(width, heads, attn_mask, droppath[i], T)
+                for i in range(layers)
+            ]
+        )
+
+    def forward(self, x: torch.Tensor):
+        if not self.use_checkpoint:
+            return self.resblocks(x)
+        else:
+            return checkpoint_sequential(self.resblocks, 3, x)
+
+
+class CrossFrameCommunicationTransformer(nn.Module):
+    def __init__(
+        self,
+        input_resolution: int,
+        patch_size: int,
+        width: int,
+        layers: int,
+        heads: int,
+        output_dim: int,
+        droppath=None,
+        T=8,
+        use_checkpoint=False,
+    ):
+        super().__init__()
+        self.input_resolution = input_resolution
+        self.output_dim = output_dim
+
+        self.conv1 = nn.Conv2d(
+            in_channels=3,
+            out_channels=width,
+            kernel_size=patch_size,
+            stride=patch_size,
+            bias=False,
+        )
+
+        scale = width ** -0.5
+        self.class_embedding = nn.Parameter(scale * torch.randn(width))
+        self.positional_embedding = nn.Parameter(
+            scale * torch.randn((input_resolution // patch_size) ** 2 + 1, width)
+        )
+        self.ln_pre = LayerNorm(width)
+
+        ## Attention Blocks
+        self.transformer = Transformer(
+            width, layers, heads, droppath=droppath, use_checkpoint=use_checkpoint, T=T,
+        )
+        self.ln_post = LayerNorm(width)
+        self.proj = nn.Parameter(scale * torch.randn(width, output_dim))
+
+    def init_weights(self):
+        self.apply(self._init_weights)
+
+    def _init_weights(self, m):
+        if isinstance(m, nn.Linear):
+            trunc_normal_(m.weight, std=0.02)
+            if isinstance(m, nn.Linear) and m.bias is not None:
+                nn.init.constant_(m.bias, 0)
+        elif isinstance(m, nn.LayerNorm):
+            nn.init.constant_(m.bias, 0)
+            nn.init.constant_(m.weight, 1.0)
+
+    def forward(self, x: torch.Tensor):
+        x = self.conv1(x)  # shape = [*, width, grid, grid]
+        x = x.reshape(x.shape[0], x.shape[1], -1)  # shape = [*, width, grid ** 2]
+        x = x.permute(0, 2, 1)  # shape = [*, grid ** 2, width]
+        x = torch.cat(
+            [
+                self.class_embedding.to(x.dtype)
+                + torch.zeros(
+                    x.shape[0], 1, x.shape[-1], dtype=x.dtype, device=x.device
+                ),
+                x,
+            ],
+            dim=1,
+        )  # shape = [*, grid ** 2 + 1, width]
+        x = x + self.positional_embedding.to(x.dtype)
+
+        x = self.ln_pre(x)
+
+        x = x.permute(1, 0, 2)
+        x = self.transformer(x)
+        x = x.permute(1, 0, 2)
+
+        cls_x = self.ln_post(x[:, 0, :])
+
+        if self.proj is not None:
+            cls_x = cls_x @ self.proj
+
+        return cls_x, x[:, 1:, :]
+
+
+class MulitHeadAttention(nn.Module):
+    def __init__(
+        self,
+        dim,
+        num_heads=8,
+        qkv_bias=False,
+        qk_scale=None,
+        attn_drop=0.0,
+        proj_drop=0.0,
+    ):
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+
+        self.scale = qk_scale or head_dim ** -0.5
+
+        self.q_proj = nn.Linear(dim, dim, bias=qkv_bias)
+        self.k_proj = nn.Linear(dim, dim, bias=qkv_bias)
+        self.v_proj = nn.Linear(dim, dim, bias=qkv_bias)
+
+        self.attn_drop = nn.Dropout(attn_drop)
+        self.proj = nn.Linear(dim, dim)
+        self.proj_drop = nn.Dropout(proj_drop)
+
+    def forward(self, q, k, v):
+        B, N, C = q.shape
+        B, M, C = k.shape
+        q = (
+            self.q_proj(q)
+            .reshape(B, N, self.num_heads, C // self.num_heads)
+            .permute(0, 2, 1, 3)
+        )
+        k = (
+            self.k_proj(k)
+            .reshape(B, M, self.num_heads, C // self.num_heads)
+            .permute(0, 2, 1, 3)
+        )
+        v = (
+            self.v_proj(v)
+            .reshape(B, M, self.num_heads, C // self.num_heads)
+            .permute(0, 2, 1, 3)
+        )
+
+        attn = (q @ k.transpose(-2, -1)) * self.scale
+        attn = attn.softmax(dim=-1)
+        attn = self.attn_drop(attn)
+
+        x = (attn @ v).transpose(1, 2).reshape(B, N, C)
+        x = self.proj(x)
+        x = self.proj_drop(x)
+        return x
+
+
+class PromptGeneratorLayer(nn.Module):
+    def __init__(
+        self, d_model, nhead, dropout=0.0,
+    ):
+        super().__init__()
+        self.cross_attn = MulitHeadAttention(d_model, nhead, proj_drop=dropout)
+
+        self.norm1 = nn.LayerNorm(d_model)
+        self.norm3 = nn.LayerNorm(d_model)
+
+        self.dropout = nn.Dropout(dropout)
+
+        self.mlp = nn.Sequential(
+            nn.Linear(d_model, d_model * 4),
+            QuickGELU(),
+            nn.Dropout(dropout),
+            nn.Linear(d_model * 4, d_model),
+        )
+
+    def forward(self, x, visual):
+        q = k = v = self.norm1(x)
+        x = x + self.cross_attn(q, visual, visual)
+        x = x + self.dropout(self.mlp(self.norm3(x)))
+        return x
+
+
+class VideoSpecificPrompt(nn.Module):
+    def __init__(
+        self, layers=2, embed_dim=512, alpha=0.1,
+    ):
+        super().__init__()
+        self.norm = nn.LayerNorm(embed_dim)
+        self.decoder = nn.ModuleList(
+            [PromptGeneratorLayer(embed_dim, embed_dim // 64) for _ in range(layers)]
+        )
+        self.alpha = nn.Parameter(torch.ones(embed_dim) * alpha)
+        self.apply(self._init_weights)
+
+    def _init_weights(self, m):
+        if isinstance(m, nn.Linear):
+            trunc_normal_(m.weight, std=0.02)
+            if isinstance(m, nn.Linear) and m.bias is not None:
+                nn.init.constant_(m.bias, 0)
+        elif isinstance(m, nn.LayerNorm):
+            nn.init.constant_(m.bias, 0)
+            nn.init.constant_(m.weight, 1.0)
+
+    def forward(self, text, visual):
+        B, N, C = visual.shape
+        visual = self.norm(visual)
+        for layer in self.decoder:
+            text = layer(text, visual)
+
+
+from collections import OrderedDict
+
+from timm.models.layers import trunc_normal_
+
+
+class ResidualAttentionBlock(nn.Module):
+    def __init__(self, d_model: int, n_head: int, attn_mask: torch.Tensor = None):
+        super().__init__()
+
+        self.attn = nn.MultiheadAttention(d_model, n_head)
+        self.ln_1 = nn.LayerNorm(d_model)
+        self.mlp = nn.Sequential(
+            OrderedDict(
+                [
+                    ("c_fc", nn.Linear(d_model, d_model * 4)),
+                    ("gelu", QuickGELU()),
+                    ("c_proj", nn.Linear(d_model * 4, d_model)),
+                ]
+            )
+        )
+        self.ln_2 = nn.LayerNorm(d_model)
+        self.attn_mask = attn_mask
+
+    def attention(self, x: torch.Tensor):
+        self.attn_mask = (
+            self.attn_mask.to(dtype=x.dtype, device=x.device)
+            if self.attn_mask is not None
+            else None
+        )
+        return self.attn(x, x, x, need_weights=False, attn_mask=self.attn_mask)[0]
+
+    def forward(self, x: torch.Tensor):
+        x = x + self.attention(self.ln_1(x))
+        x = x + self.mlp(self.ln_2(x))
+        return x
+
+
+class MultiframeIntegrationTransformer(nn.Module):
+    def __init__(
+        self, T, embed_dim=512, layers=1,
+    ):
+        super().__init__()
+        self.T = T
+        transformer_heads = embed_dim // 64
+        self.positional_embedding = nn.Parameter(torch.empty(1, T, embed_dim))
+        trunc_normal_(self.positional_embedding, std=0.02)
+        self.resblocks = nn.Sequential(
+            *[
+                ResidualAttentionBlock(d_model=embed_dim, n_head=transformer_heads)
+                for _ in range(layers)
+            ]
+        )
+
+        self.apply(self._init_weights)
+
+    def _init_weights(self, m):
+        if isinstance(m, (nn.Linear,)):
+            trunc_normal_(m.weight, std=0.02)
+            if m.bias is not None:
+                nn.init.zeros_(m.bias)
+        elif isinstance(m, nn.LayerNorm):
+            nn.init.zeros_(m.bias)
+            nn.init.ones_(m.weight)
+
+    def forward(self, x):
+        ori_x = x
+        x = x + self.positional_embedding
+        x = x.permute(1, 0, 2)
+        x = self.resblocks(x)
+        x = x.permute(1, 0, 2)
+        x = x.type(ori_x.dtype) + ori_x
+
+        return x.mean(dim=1, keepdim=False)
+
+
+class XCLIP(CLIP):
+    def __init__(
+        self,
+        embed_dim: int,
+        # vision
+        image_resolution: int,
+        vision_layers: Union[Tuple[int, int, int, int], int],
+        vision_width: int,
+        vision_patch_size: int,
+        # text
+        context_length: int,
+        vocab_size: int,
+        transformer_width: int,
+        transformer_heads: int,
+        transformer_layers: int,
+        # video
+        T=8,
+        droppath=0.0,
+        mit_layers=1,
+        # prompt
+        prompts_alpha=1e-4,
+        prompts_layers=1,
+        # other
+        use_cache=True,
+        use_checkpoint=False,
+    ):
+        super().__init__(
+            embed_dim,
+            image_resolution,
+            vision_layers,
+            vision_width,
+            vision_patch_size,
+            context_length,
+            vocab_size,
+            transformer_width,
+            transformer_heads,
+            transformer_layers,
+        )
+
+        self.prompts_generator = VideoSpecificPrompt(
+            layers=prompts_layers, embed_dim=embed_dim, alpha=prompts_alpha,
+        )
+        self.use_cache = use_cache
+        self.mit = MultiframeIntegrationTransformer(
+            T=T, embed_dim=embed_dim, layers=mit_layers,
+        )
+
+        dpr = (
+            [x.item() for x in torch.linspace(0, droppath, vision_layers)]
+            if droppath > 0.0
+            else None
+        )
+
+        vision_heads = vision_width // 64
+        self.visual = CrossFrameCommunicationTransformer(
+            input_resolution=image_resolution,
+            patch_size=vision_patch_size,
+            width=vision_width,
+            layers=vision_layers,
+            heads=vision_heads,
+            output_dim=embed_dim,
+            droppath=dpr,
+            T=T,
+            use_checkpoint=use_checkpoint,
+        )
+
+        self.transformer = Transformer(
+            width=transformer_width,
+            layers=transformer_layers,
+            heads=transformer_heads,
+            attn_mask=self.build_attention_mask(),
+        )
+        self.vocab_size = vocab_size
+        self.token_embedding = nn.Embedding(vocab_size, transformer_width)
+        self.positional_embedding = nn.Parameter(
+            torch.empty(self.context_length, transformer_width)
+        )
+        self.ln_final = LayerNorm(transformer_width)
+        self.text_projection = nn.Parameter(torch.empty(transformer_width, embed_dim))
+        self.logit_scale = nn.Parameter(torch.ones([]) * np.log(1 / 0.07))
+
+        self.cache_text_features = None
+        self.prompts_visual_ln = LayerNorm(vision_width)
+        self.prompts_visual_proj = nn.Parameter(torch.randn(vision_width, embed_dim))
+
+        self.initialize_parameters()
+
+    @torch.jit.ignore
+    def no_weight_decay_keywords(self):
+        return {"positional_embedding"}
+
+    def encode_image(self, image):
+        return self.visual(image)
+
+    def encode_text(self, text):
+        x = self.token_embedding(text)
+        eos_indx = text.argmax(dim=-1)
+        K, N1, C = x.shape
+
+        x = x + self.positional_embedding
+        x = x.permute(1, 0, 2)  # NLD -> LND
+        x = self.transformer(x)
+        x = x.permute(1, 0, 2)  # LND -> NLD
+        x = self.ln_final(x)
+        # x.shape = [batch_size, n_ctx, transformer.width]
+        # take features from the eot embedding (eot_token is the highest number in each sequence)
+        x = x[torch.arange(x.shape[0]), eos_indx] @ self.text_projection
+        x = x.reshape(K, -1)
+        return x
+
+    def encode_video(self, image):
+        b, t, c, h, w = image.size()
+        image = image.reshape(-1, c, h, w)
+
+        cls_features, img_features = self.encode_image(image)
+        img_features = self.prompts_visual_ln(img_features)
+        img_features = img_features @ self.prompts_visual_proj
+
+        cls_features = cls_features.view(b, t, -1)
+        img_features = img_features.view(b, t, -1, cls_features.shape[-1])
+
+        video_features = self.mit(cls_features)
+
+        return video_features, img_features
+
+    def forward(self, image, **kwargs):
+        image = rearrange(image, "b c t h w -> b t c h w")
+        video_features, _ = self.encode_video(image)
+        return video_features.reshape(*video_features.shape, 1, 1, 1)
+
+    def cache_text(self, text):
+        self.eval()
+        with torch.no_grad():
+            if self.cache_text_features is None:
+                self.cache_text_features = self.encode_text(text)
+        self.train()
+        return self.cache_text_features
+
+    def forward_original(self, image, text):
+        b = image.shape[0]
+        video_features, img_features = self.encode_video(image)
+        img_features = img_features.mean(dim=1, keepdim=False)
+
+        if self.use_cache:
+            text_features = self.cache_text(text)
+        else:
+            text_features = self.encode_text(text)
+
+        text_features = text_features.unsqueeze(0).expand(b, -1, -1)
+        text_features = text_features + self.prompts_generator(
+            text_features, img_features
+        )
+
+        video_features = video_features / video_features.norm(dim=-1, keepdim=True)
+        text_features = text_features / text_features.norm(dim=-1, keepdim=True)
+        logit_scale = self.logit_scale.exp()
+        logits = torch.einsum("bd,bkd->bk", video_features, logit_scale * text_features)
+
+        return logits
+
+
+def build_x_clip_model(
+    pretrained_path="./pretrained_weights/k400_32_8.pth",
+    droppath=0.0,
+    use_checkpoint=False,
+    logger=None,
+    prompts_alpha=1e-1,
+    prompts_layers=2,
+    use_cache=True,
+    mit_layers=4,
+    **kwargs,
+):
+    state_dict = torch.load(pretrained_path, map_location="cpu")["model"]
+    T = int(pretrained_path.split("_")[-1].split(".")[0])
+    print(T)
+    vit = "visual.proj" in state_dict
+
+    if vit:
+        vision_width = state_dict["visual.conv1.weight"].shape[0]
+        vision_layers = len(
+            [
+                k
+                for k in state_dict.keys()
+                if k.startswith("visual.") and k.endswith(".attn.in_proj_weight")
+            ]
+        )
+        vision_patch_size = state_dict["visual.conv1.weight"].shape[-1]
+        grid_size = round(
+            (state_dict["visual.positional_embedding"].shape[0] - 1) ** 0.5
+        )
+        image_resolution = vision_patch_size * grid_size
+    else:
+        counts: list = [
+            len(
+                set(
+                    k.split(".")[2]
+                    for k in state_dict
+                    if k.startswith(f"visual.layer{b}")
+                )
+            )
+            for b in [1, 2, 3, 4]
+        ]
+        vision_layers = tuple(counts)
+
+        vision_width = state_dict["visual.layer1.0.conv1.weight"].shape[0]
+        output_width = round(
+            (state_dict["visual.attnpool.positional_embedding"].shape[0] - 1) ** 0.5
+        )
+        vision_patch_size = None
+        assert (
+            output_width ** 2 + 1
+            == state_dict["visual.attnpool.positional_embedding"].shape[0]
+        )
+        image_resolution = output_width * 32
+
+    embed_dim = state_dict["text_projection"].shape[1]
+    context_length = state_dict["positional_embedding"].shape[0]
+    vocab_size = state_dict["token_embedding.weight"].shape[0]
+    transformer_width = state_dict["ln_final.weight"].shape[0]
+    transformer_heads = transformer_width // 64
+    transformer_layers = len(
+        set(
+            k.split(".")[2]
+            for k in state_dict
+            if k.startswith(f"transformer.resblocks")
+        )
+    )
+
+    model = XCLIP(
+        embed_dim,
+        image_resolution,
+        vision_layers,
+        vision_width,
+        vision_patch_size,
+        context_length,
+        vocab_size,
+        transformer_width,
+        transformer_heads,
+        transformer_layers,
+        T=T,
+        droppath=droppath,
+        mit_layers=mit_layers,
+        prompts_alpha=prompts_alpha,
+        prompts_layers=prompts_layers,
+        use_checkpoint=use_checkpoint,
+        use_cache=use_cache,
+    )
+
+    for key in ["input_resolution", "context_length", "vocab_size"]:
+        if key in state_dict:
+            del state_dict[key]
+
+    msg = model.load_state_dict(state_dict, strict=False)
+
+    return model.eval()

cover/version.py

@@ -0,0 +1,16 @@
+__version__ = "1.0.0"
+
+
+def parse_version_info(version_str):
+    version_info = []
+    for x in version_str.split("."):
+        if x.isdigit():
+            version_info.append(int(x))
+        elif x.find("rc") != -1:
+            patch_version = x.split("rc")
+            version_info.append(int(patch_version[0]))
+            version_info.append(f"rc{patch_version[1]}")
+    return tuple(version_info)
+
+
+version_info = parse_version_info(__version__)

evaluate_a_set_of_videos.py

@@ -0,0 +1,119 @@
+import torch
+
+import argparse
+import os
+import pickle as pkl
+
+import decord
+import numpy as np
+import yaml
+from tqdm import tqdm
+
+from cover.datasets import (
+    UnifiedFrameSampler,
+    ViewDecompositionDataset,
+    spatial_temporal_view_decomposition,
+)
+from cover.models import COVER
+
+mean, std = (
+    torch.FloatTensor([123.675, 116.28, 103.53]),
+    torch.FloatTensor([58.395, 57.12, 57.375]),
+)
+
+mean_clip, std_clip = (
+    torch.FloatTensor([122.77, 116.75, 104.09]),
+    torch.FloatTensor([68.50, 66.63, 70.32])
+)
+
+def fuse_results(results: list):
+    x = (results[0] + results[1] + results[2])
+    return {
+        "semantic" : results[0],
+        "technical": results[1],
+        "aesthetic": results[2],
+        "overall"  : x,
+    }
+
+def parse_args():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("-o", "--opt"   , type=str, default="./cover.yml", help="the option file")
+    parser.add_argument('-d', "--device", type=str, default="cuda"       , help='CUDA device id')
+    parser.add_argument("-i", "--input_video_dir", type=str, default="./demo", help="the input video dir")
+    parser.add_argument(      "--output", type=str, default="./demo.csv" , help='output file to store predict mos value')
+    args = parser.parse_args()
+    return args
+
+
+if __name__ == "__main__":
+
+    args = parse_args()
+
+    with open(args.opt, "r") as f:
+        opt = yaml.safe_load(f)
+
+    ### Load COVER
+    evaluator = COVER(**opt["model"]["args"]).to(args.device)
+    state_dict = torch.load(opt["test_load_path"], map_location=args.device)
+    
+    # set strict=False here to avoid error of missing
+    # weight of prompt_learner in clip-iqa+, cross-gate
+    evaluator.load_state_dict(state_dict['state_dict'], strict=False)
+
+
+    video_paths = []
+    all_results = {}
+
+    with open(args.output, "w") as w:
+        w.write(f"path, semantic score, technical score, aesthetic score, overall/final score\n")
+
+    dopt = opt["data"]["val-l1080p"]["args"]
+
+    dopt["anno_file"] = None
+    dopt["data_prefix"] = args.input_video_dir
+
+    dataset = ViewDecompositionDataset(dopt)
+
+    dataloader = torch.utils.data.DataLoader(
+        dataset, batch_size=1, num_workers=opt["num_workers"], pin_memory=True,
+    )
+
+    sample_types = ["semantic", "technical", "aesthetic"]
+
+    for i, data in enumerate(tqdm(dataloader, desc="Testing")):
+        if len(data.keys()) == 1:
+            ##  failed data
+            continue
+
+        video = {}
+        for key in sample_types:
+            if key in data:
+                video[key] = data[key].to(args.device)
+                b, c, t, h, w = video[key].shape
+                video[key] = (
+                    video[key]
+                    .reshape(
+                        b, c, data["num_clips"][key], t // data["num_clips"][key], h, w
+                    )
+                    .permute(0, 2, 1, 3, 4, 5)
+                    .reshape(
+                        b * data["num_clips"][key], c, t // data["num_clips"][key], h, w
+                    )
+                )
+    
+        with torch.no_grad():
+            results = evaluator(video, reduce_scores=False)
+            results = [np.mean(l.cpu().numpy()) for l in results]
+
+        rescaled_results = fuse_results(results)
+        # all_results[data["name"][0]] = rescaled_results
+
+        # with open(
+        #    f"cover_predictions/val-custom_{args.input_video_dir.split('/')[-1]}.pkl", "wb"
+        # ) as wf:
+        # pkl.dump(all_results, wf)
+        
+        with open(args.output, "a") as w:
+            w.write(
+                f'{data["name"][0].split("/")[-1]},{rescaled_results["semantic"]:4f},{rescaled_results["technical"]:4f},{rescaled_results["aesthetic"]:4f},{rescaled_results["overall"]:4f}\n'
+            )

evaluate_one_dataset.py

@@ -0,0 +1,190 @@
+import argparse
+import os
+import csv
+
+import torch
+
+import pandas as pd
+import numpy as np
+import pickle as pkl
+import decord
+import yaml
+
+from scipy import stats
+from sklearn.metrics import mean_squared_error
+from scipy.optimize import curve_fit
+from cover.datasets import UnifiedFrameSampler, spatial_temporal_view_decomposition
+from cover.models import COVER
+
+# use case
+# python evaluate_on_ytugc.py -o cover.yml -d cuda:3 --output result.csv -uh 0
+
+def save_to_csv(video_name, pre_smos, pre_tmos, pre_amos, pre_overall, filename):
+    combined_data = list(zip(video_name, pre_smos, pre_tmos, pre_amos, pre_overall))
+
+    with open(filename, 'w', newline='') as csvfile:
+        writer = csv.writer(csvfile)
+        writer.writerow(['Video', 'semantic score', 'technical score', 'aesthetic score', 'overall/final score'])
+        writer.writerows(combined_data)
+
+mean_cover, std_cover = (
+    torch.FloatTensor([123.675, 116.28, 103.53]),
+    torch.FloatTensor([58.395, 57.12, 57.375]),
+)
+
+mean_clip, std_clip = (
+    torch.FloatTensor([122.77, 116.75, 104.09]),
+    torch.FloatTensor([68.50, 66.63, 70.32])
+)
+
+def fuse_results(results: list):
+    x = (results[0] + results[1] + results[2])
+    return {
+        "semantic" : results[0],
+        "technical": results[1],
+        "aesthetic": results[2],
+        "overall"  : x,
+    }
+
+
+def gaussian_rescale(pr):
+    # The results should follow N(0,1)
+    pr = (pr - np.mean(pr)) / np.std(pr)
+    return pr
+
+ 
+def uniform_rescale(pr):
+    # The result scores should follow U(0,1)
+    return np.arange(len(pr))[np.argsort(pr).argsort()] / len(pr)
+
+
+def parse_args():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("-o", "--opt"   , type=str, default="./cover.yml", help="the option file")
+    parser.add_argument('-d', "--device", type=str, default="cuda:0"     , help='CUDA device id')
+    parser.add_argument("-t", "--target_set", type=str, default="val-ytugc", help="target_set")
+    parser.add_argument(      "--output", type=str, default="ytugc.csv" , help='output file to store predict mos value')
+    args = parser.parse_args()
+    return args
+
+
+def logistic_func(X, bayta1, bayta2, bayta3, bayta4):
+    # 4-parameter logistic function
+    logisticPart = 1 + np.exp(np.negative(np.divide(X - bayta3, np.abs(bayta4))))
+    yhat = bayta2 + np.divide(bayta1 - bayta2, logisticPart)
+    return yhat
+
+
+if __name__ == '__main__':
+    args = parse_args()
+
+    with open(args.opt, "r") as f:
+        opt = yaml.safe_load(f)
+    
+    ### Load COVER
+    evaluator = COVER(**opt["model"]["args"]).to(args.device)
+    state_dict = torch.load(opt["test_load_path"], map_location=args.device)
+
+    # set strict=False here to avoid error of missing
+    # weight of prompt_learner in clip-iqa+, cross-gate
+    evaluator.load_state_dict(state_dict['state_dict'], strict=False)
+
+    dopt = opt["data"][args.target_set]["args"]
+    temporal_samplers = {}
+    for stype, sopt in dopt["sample_types"].items():
+        temporal_samplers[stype] = UnifiedFrameSampler(
+            sopt["clip_len"] // sopt["t_frag"],
+            sopt["t_frag"],
+            sopt["frame_interval"],
+            sopt["num_clips"],
+        )
+
+    if args.target_set == 'val-livevqc':
+        videos_dir = './datasets/LIVE_VQC/Video/'
+        datainfo = './datasets/LIVE_VQC/metainfo/LIVE_VQC_metadata.csv'
+        df = pd.read_csv(datainfo)
+        files = df['File'].tolist()
+        mos = df['MOS'].tolist()
+    elif args.target_set == 'val-kv1k':
+        videos_dir = './datasets/KoNViD/KoNViD_1k_videos/'
+        datainfo = './datasets/KoNViD/metainfo/KoNVid_metadata.csv'
+        df = pd.read_csv(datainfo)
+        files = df['Filename'].tolist()
+        files = [str(file) + '.mp4' for file in files]
+        mos = df['MOS'].tolist()
+    elif args.target_set == 'val-ytugc':
+        videos_dir = './datasets/YouTubeUGC/'
+        datainfo = './datasets/YouTubeUGC/../meta_info/Youtube-UGC_metadata.csv'
+        df = pd.read_csv(datainfo)
+        files = df['filename'].tolist()
+        mos = df['MOSFull'].tolist()
+        files = [str(file) + '_crf_10_ss_00_t_20.0.mp4' for file in files]
+    else:
+        print("unsupported video dataset for evaluation")
+        assert(0)
+
+    print(len(files))
+
+    pure_name_list = []
+    pre_overall = np.zeros(len(mos))
+    pre_smos = np.zeros(len(mos))
+    pre_tmos = np.zeros(len(mos))
+    pre_amos = np.zeros(len(mos))
+    gt_mos = np.array(mos)
+    count = 0
+
+    for vi in range(len(mos)):
+        video = files[vi]
+        pure_name = os.path.splitext(video)[0]
+        video_path = os.path.join(videos_dir, video)
+
+        views, _ = spatial_temporal_view_decomposition(
+            video_path, dopt["sample_types"], temporal_samplers
+        )
+        
+        for k, v in views.items():
+            num_clips = dopt["sample_types"][k].get("num_clips", 1)
+            if k == 'technical' or k == 'aesthetic':
+                views[k] = (
+                    ((v.permute(1, 2, 3, 0) - mean_cover) / std_cover)
+                    .permute(3, 0, 1, 2)
+                    .reshape(v.shape[0], num_clips, -1, *v.shape[2:])
+                    .transpose(0, 1)
+                    .to(args.device)
+                )
+            elif k == 'semantic':
+                views[k] = (
+                    ((v.permute(1, 2, 3, 0) - mean_clip) / std_clip)
+                    .permute(3, 0, 1, 2)
+                    .reshape(v.shape[0], num_clips, -1, *v.shape[2:])
+                    .transpose(0, 1)
+                    .to(args.device)
+                )
+
+        results = [r.mean().item() for r in evaluator(views)]
+
+    
+        pre_overall[count] = fuse_results(results)['overall']
+        pre_smos[count] = results[0]
+        pre_tmos[count] = results[1]
+        pre_amos[count] = results[2]
+        pure_name_list.append(pure_name)
+        print("Process ", video, ", predicted quality score is ", pre_overall[count])
+        count += 1
+
+
+    SROCC = stats.spearmanr(pre_overall, gt_mos)[0]
+    KROCC = stats.stats.kendalltau(pre_overall, gt_mos)[0]
+
+    # logistic regression btw y_pred & y
+    beta_init = [np.max(gt_mos), np.min(gt_mos), np.mean(pre_overall), 0.5]
+    popt, _ = curve_fit(logistic_func, pre_overall, gt_mos, p0=beta_init, maxfev=int(1e8))
+    pre_overall_logistic = logistic_func(pre_overall, *popt)
+
+    PLCC = stats.pearsonr(gt_mos, pre_overall_logistic)[0]
+    RMSE = np.sqrt(mean_squared_error(gt_mos, pre_overall_logistic))
+
+    print("Test results: SROCC={:.4f}, KROCC={:.4f}, PLCC={:.4f}, RMSE={:.4f}"
+          .format(SROCC, KROCC, PLCC, RMSE))
+
+    save_to_csv(pure_name_list, pre_smos, pre_tmos, pre_amos, pre_overall, args.output)

evaluate_one_video.py

@@ -0,0 +1,105 @@
+import torch
+
+import argparse
+import pickle as pkl
+
+import decord
+from decord import VideoReader
+import numpy as np
+import yaml
+
+from cover.datasets import UnifiedFrameSampler, spatial_temporal_view_decomposition
+from cover.models import COVER
+
+mean, std = (
+    torch.FloatTensor([123.675, 116.28, 103.53]),
+    torch.FloatTensor([58.395, 57.12, 57.375]),
+)
+
+mean_clip, std_clip = (
+    torch.FloatTensor([122.77, 116.75, 104.09]),
+    torch.FloatTensor([68.50, 66.63, 70.32])
+)
+
+def fuse_results(results: list):
+    x = (results[0] + results[1] + results[2])
+    return {
+        "semantic" : results[0],
+        "technical": results[1],
+        "aesthetic": results[2],
+        "overall"  : x,
+    }
+
+
+def parse_args():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("-o", "--opt"   , type=str, default="./cover.yml", help="the option file")
+    parser.add_argument("--video_path", type=str, default="./demo/video_1.mp4" , help='output file to store predict mos value')
+    args = parser.parse_args()
+    return args
+
+if __name__ == "__main__":
+
+    args = parse_args()
+
+    """
+    BASIC SETTINGS
+    """
+    torch.cuda.current_device()
+    torch.cuda.empty_cache()
+    torch.backends.cudnn.benchmark = True
+    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+    with open(args.opt, "r") as f:
+       opt = yaml.safe_load(f)
+    
+    dopt = opt["data"]["val-ytugc"]["args"]
+    temporal_samplers = {}
+    for stype, sopt in dopt["sample_types"].items():
+        temporal_samplers[stype] = UnifiedFrameSampler(
+            sopt["clip_len"] // sopt["t_frag"],
+            sopt["t_frag"],
+            sopt["frame_interval"],
+            sopt["num_clips"],
+        )
+
+    """
+    LOAD MODEL
+    """    
+    evaluator = COVER(**opt["model"]["args"]).to(device)
+    state_dict = torch.load(opt["test_load_path"], map_location=device)
+    
+    # set strict=False here to avoid error of missing
+    # weight of prompt_learner in clip-iqa+, cross-gate
+    evaluator.load_state_dict(state_dict['state_dict'], strict=False)
+
+    """
+    TESTING
+    """
+    views, _ = spatial_temporal_view_decomposition(
+        args.video_path, dopt["sample_types"], temporal_samplers
+    )
+
+    for k, v in views.items():
+        num_clips = dopt["sample_types"][k].get("num_clips", 1)
+        if k == 'technical' or k == 'aesthetic':
+            views[k] = (
+                ((v.permute(1, 2, 3, 0) - mean) / std)
+                .permute(3, 0, 1, 2)
+                .reshape(v.shape[0], num_clips, -1, *v.shape[2:])
+                .transpose(0, 1)
+                .to(device)
+            )
+        elif k == 'semantic':
+            views[k] = (
+                ((v.permute(1, 2, 3, 0) - mean_clip) / std_clip)
+                .permute(3, 0, 1, 2)
+                .reshape(v.shape[0], num_clips, -1, *v.shape[2:])
+                .transpose(0, 1)
+                .to(device)
+            )
+
+    results = [r.mean().item() for r in evaluator(views)]
+    pred_score = fuse_results(results)
+    print(f"path, semantic score, technical score, aesthetic score, overall/final score")
+    print(f'{args.video_path.split("/")[-1]},{pred_score["semantic"]:4f},{pred_score["technical"]:4f},{pred_score["aesthetic"]:4f},{pred_score["overall"]:4f}')
+

examplar_data_labels/CVD2014/labels.txt

@@ -0,0 +1,234 @@
+Test3/City/Test03_City_D12.avi, 14.948471, 29.099967117037156, 20.65
+Test5/City/Test05_City_D05.avi, 15.0, 24.0, 42.92
+Test3/Talking_head/Test03_Talking_Head_D10.avi, 16.4, 25.0, 53.33
+Test5/Television/Test05_Television_D03.avi, 21.04, 25.0, 63.93
+Test5/City/Test05_City_D04.avi, 15.014389, 23.977, 63.85
+Test4/City/Test04_City_D01.avi, 15.015015, 29.97000002997, 61.85
+Test3/Talking_head/Test03_Talking_Head_D08.avi, 16.308336, 24.649970420035498, 46.1
+Test2/City/Test02_City_D05.avi, 15.755979, 29.06833092550659, 45.91
+Test4/Talking_head/Test04_Talking_Head_D14.avi, 14.983296, 23.96001552609006, 52.76
+Test3/Talking_head/Test03_Talking_Head_D09.avi, 16.36177, 29.520033770918634, 37.85
+Test4/Newspaper/Test04_Newspaper_D08.avi, 14.98327, 14.950007176003444, 29.73
+Test1/Traffic/Test01_Traffic_D01.avi, 23.166435, 30.00030000300003, 67.74
+Test6/Talking_head/Test06_Talking_head_D12.avi, 15.011408999999999, 24.981, 57.39
+Test3/City/Test03_City_D02.avi, 15.048382, 29.97000002997, 16.48
+Test5/Television/Test05_Television_D11.avi, 20.488353, 24.941, 23.3
+Test3/City/Test03_City_D13.avi, 14.976324, 29.58002283577763, 32.65
+Test2/Talking_head/Test02_Talking_Head_D07.avi, 16.832148, 29.526832509042592, 36.58
+Test1/Talking_head/Test01_Talking_Head_D06.avi, 16.638224, 27.52697643690817, 38.22
+Test5/Television/Test05_Television_D06.avi, 20.615546, 17.123, 33.32
+Test3/Newspaper/Test03_Newspaper_D05.avi, 15.015015, 29.97000002997, 40.1
+Test5/Talking_head/Test05_Talking_Head_D05.avi, 15.0, 24.0, 69.78
+Test5/City/Test05_City_D10.avi, 14.999302, 28.668, 65.29
+Test6/Talking_head/Test06_Talking_head_D01.avi, 15.0, 30.8, 14.9
+Test3/Talking_head/Test03_Talking_Head_D01.avi, 15.549999999999999, 20.0, 39.03
+Test2/Traffic/Test02_Traffic_D03.avi, 25.726547, 14.96508645330444, 21.78
+Test3/Talking_head/Test03_Talking_Head_D12.avi, 16.288678, 29.099967117037156, 20.86
+Test1/Talking_head/Test01_Talking_Head_D03.avi, 17.260519, 14.715663941325705, 31.98
+Test3/City/Test03_City_D11.avi, 15.204692999999999, 22.229978014551744, 15.47
+Test1/Talking_head/Test01_Talking_Head_D09.avi, 16.12, 25.0, 64.59
+Test6/City/Test06_City_D01.avi, 15.0, 30.8, 17.23
+Test5/Talking_head/Test05_Talking_Head_D07.avi, 14.979391, 19.894, 73.71
+Test3/Newspaper/Test03_Newspaper_D06.avi, 14.966444, 14.89999865900012, 31.2
+Test1/Traffic/Test01_Traffic_D08.avi, 24.119999999999997, 25.0, 63.4
+Test2/Traffic/Test02_Traffic_D07.avi, 25.966395, 29.461155466517397, 61.43
+Test4/Newspaper/Test04_Newspaper_D07.avi, 15.023477999999999, 14.909995810291177, 32.42
+Test3/Talking_head/Test03_Talking_Head_D11.avi, 16.115344, 23.5800116956858, 38.37
+Test4/Talking_head/Test04_Talking_Head_D13.avi, 13.04638, 29.97000002997, 59.4
+Test1/Traffic/Test01_Traffic_D07.avi, 22.068345, 29.453953162323682, 65.53
+Test3/Talking_head/Test03_Talking_Head_D06.avi, 15.234900999999999, 14.89999865900012, 28.2
+Test5/Talking_head/Test05_Talking_Head_D03.avi, 15.0, 25.0, 83.25
+Test2/Talking_head/Test02_Talking_Head_D05.avi, 17.334484, 28.728862739239606, 46.3
+Test6/Television/Test06_Television_D11.avi, 22.015213, 29.843, 23.8
+Test4/Newspaper/Test04_Newspaper_D09.avi, 15.013294, 30.040043377822638, 87.46
+Test4/City/Test04_City_D02.avi, 15.0, 25.0, 70.56
+Test6/City/Test06_City_D07.avi, 15.009443999999998, 29.648, 58.78
+Test1/Traffic/Test01_Traffic_D03.avi, 26.332442, 14.848603637313946, 27.76
+Test6/Talking_head/Test06_Talking_head_D10.avi, 14.997193, 24.938, 23.31
+Test5/City/Test05_City_D08.avi, 15.000497, 30.199, 63.76
+Test4/Talking_head/Test04_Talking_Head_D01.avi, 14.839504, 23.99002015161693, 51.43
+Test1/Talking_head/Test01_Talking_Head_D05.avi, 17.148035999999998, 14.753875843184005, 35.1
+Test6/City/Test06_City_D12.avi, 15.009005, 24.985, 53.03
+Test6/Talking_head/Test06_Talking_head_D04.avi, 15.0, 25.0, 76.79
+Test6/Television/Test06_Television_D03.avi, 21.997249999999998, 24.003, 28.29
+Test3/City/Test03_City_D04.avi, 15.006241, 21.124543973906963, 13.96
+Test3/Newspaper/Test03_Newspaper_D07.avi, 15.003326999999999, 30.059999759520004, 49.85
+Test1/City/Test01_City_D04.avi, 11.498643, 30.003540417769297, 30.23
+Test5/City/Test05_City_D03.avi, 15.0, 25.0, 68.62
+Test5/Television/Test05_Television_D12.avi, 20.735758999999998, 24.981, 35.12
+Test3/Newspaper/Test03_Newspaper_D10.avi, 15.0, 25.0, 49.95
+Test2/Talking_head/Test02_Talking_Head_D10.avi, 16.846667999999998, 29.79817694753435, 33.8
+Test5/City/Test05_City_D09.avi, 15.0, 30.0, 79.61
+Test3/City/Test03_City_D10.avi, 15.008216, 24.320012451846374, 50.12
+Test5/Talking_head/Test05_Talking_Head_D16.avi, 15.016328999999999, 29.701, 72.57
+Test2/Talking_head/Test02_Talking_Head_D06.avi, 16.845816, 23.151149223047433, 24.35
+Test6/Talking_head/Test06_Talking_head_D09.avi, 15.033011, 12.173210172908277, 0.58
+Test6/Television/Test06_Television_D02.avi, 21.990837, 24.01, 28.81
+Test4/Talking_head/Test04_Talking_Head_D09.avi, 13.48391, 30.109961579689024, 75.11
+Test6/City/Test06_City_D05.avi, 15.005583, 28.656, 61.12
+Test2/City/Test02_City_D04.avi, 16.499721, 14.848735927110525, 9.25
+Test5/City/Test05_City_D11.avi, 15.007503999999999, 29.985, 73.55
+Test5/Television/Test05_Television_D15.avi, 20.688726, 29.678, 36.1
+Test4/Talking_head/Test04_Talking_Head_D05.avi, 14.347681, 29.97000002997, 93.38
+Test6/City/Test06_City_D08.avi, 15.0, 30.0, 72.68
+Test4/Newspaper/Test04_Newspaper_D14.avi, 15.012504, 23.98001026344439, 41.92
+Test1/City/Test01_City_D06.avi, 11.363332999999999, 28.600764784450337, 43.78
+Test5/Talking_head/Test05_Talking_Head_D15.avi, 14.983148, 29.967, 63.92
+Test6/City/Test06_City_D11.avi, 15.011896, 29.843, 62.03
+Test1/City/Test01_City_D05.avi, 11.708831, 14.689766814641585, 25.08
+Test2/Talking_head/Test02_Talking_Head_D09.avi, 16.919999999999998, 25.0, 64.84
+Test4/Talking_head/Test04_Talking_Head_D02.avi, 15.52, 25.0, 83.6
+Test4/Newspaper/Test04_Newspaper_D06.avi, 15.027104, 12.910005641672466, 5.08
+Test6/City/Test06_City_D06.avi, 14.983493, 23.626, 55.78
+Test6/Talking_head/Test06_Talking_head_D11.avi, 14.990084999999999, 29.753, 42.32
+Test3/Talking_head/Test03_Talking_Head_D03.avi, 15.799983999999998, 30.00003000003, 67.38
+Test5/City/Test05_City_D12.avi, 15.010207, 24.983, 64.24
+Test4/Newspaper/Test04_Newspaper_D13.avi, 15.015015, 29.97000002997, 64.41
+Test6/Television/Test06_Television_D05.avi, 22.01717, 17.123, 27.85
+Test1/Talking_head/Test01_Talking_Head_D02.avi, 17.028, 15.151515151515152, 31.65
+Test1/Talking_head/Test01_Talking_Head_D01.avi, 16.166505, 30.00030000300003, 69.37
+Test3/Newspaper/Test03_Newspaper_D13.avi, 15.11155, 29.58002283577763, 38.65
+Test5/City/Test05_City_D01.avi, 15.000487, 30.799, 26.07
+Test2/Talking_head/Test02_Talking_Head_D01.avi, 16.966497, 30.00030000300003, 64.68
+Test4/Newspaper/Test04_Newspaper_D11.avi, 14.985992, 14.280002284800366, 4.14
+Test3/Newspaper/Test03_Newspaper_D08.avi, 14.945312, 24.690016838591482, 47.8
+Test3/Newspaper/Test03_Newspaper_D09.avi, 15.013591, 29.43999246336193, 47.22
+Test6/Television/Test06_Television_D08.avi, 22.000733, 29.999, 23.61
+Test2/Traffic/Test02_Traffic_D05.avi, 26.822181999999998, 28.670300149372263, 30.89
+Test4/City/Test04_City_D05.avi, 15.015015, 29.97000002997, 65.86
+Test4/Talking_head/Test04_Talking_Head_D11.avi, 14.856345, 14.269996903410672, 4.5
+Test2/Traffic/Test02_Traffic_D06.avi, 25.941309999999998, 24.169943708201103, 34.66
+Test2/City/Test02_City_D06.avi, 14.589739, 24.949041582567606, 38.99
+Test5/Talking_head/Test05_Talking_Head_D04.avi, 15.015015, 23.976, 80.15
+Test5/Television/Test05_Television_D13.avi, 20.279189, 25.001, 27.91
+Test6/Television/Test06_Television_D06.avi, 22.01384, 19.942, 33.68
+Test5/Television/Test05_Television_D05.avi, 21.375, 24.0, 48.27
+Test5/Television/Test05_Television_D01.avi, 20.975388, 30.798, 11.45
+Test6/Talking_head/Test06_Talking_head_D03.avi, 14.997499999999999, 24.004, 53.44
+Test3/City/Test03_City_D08.avi, 14.033662, 24.369975215735206, 30.01
+Test1/City/Test01_City_D01.avi, 12.133212, 30.00030000300003, 67.29
+Test4/City/Test04_City_D08.avi, 15.030068, 14.969992649733609, 26.17
+Test1/Talking_head/Test01_Talking_Head_D08.avi, 16.08, 25.0, 53.69
+Test2/Traffic/Test02_Traffic_D02.avi, 20.776552, 29.985726794046034, 25.39
+Test2/Traffic/Test02_Traffic_D01.avi, 26.866398, 30.00030000300003, 69.62
+Test6/Television/Test06_Television_D09.avi, 22.022837, 10.670741467141585, 7.23
+Test5/Television/Test05_Television_D07.avi, 20.409357, 17.1, 43.05
+Test4/City/Test04_City_D03.avi, 15.0, 25.0, 68.6
+Test2/City/Test02_City_D08.avi, 15.837328, 29.73986539736921, 36.88
+Test5/Talking_head/Test05_Talking_Head_D02.avi, 14.993753, 24.01, 65.4
+Test2/City/Test02_City_D10.avi, 16.454064, 29.415225320625957, 0.21
+Test4/Newspaper/Test04_Newspaper_D01.avi, 15.015015, 29.97000002997, 70.38
+Test4/City/Test04_City_D10.avi, 15.018037999999999, 24.9699736067379, 38.64
+Test2/Talking_head/Test02_Talking_Head_D02.avi, 16.564215, 30.00444065721727, 22.64
+Test2/Traffic/Test02_Traffic_D08.avi, 25.939173, 29.87759150759339, 29.35
+Test2/City/Test02_City_D01.avi, 15.333179999999999, 30.00030000300003, 53.35
+Test6/Talking_head/Test06_Talking_head_D08.avi, 15.0, 30.0, 62.56
+Test3/Newspaper/Test03_Newspaper_D11.avi, 15.014636999999999, 23.910002749650317, 46.51
+Test6/Talking_head/Test06_Talking_head_D07.avi, 15.001681999999999, 14.865, 32.99
+Test4/Newspaper/Test04_Newspaper_D10.avi, 14.983972999999999, 24.9600015974401, 52.99
+Test6/City/Test06_City_D10.avi, 14.999630999999999, 27.134, 23.58
+Test6/City/Test06_City_D13.avi, 15.015452, 27.505, 64.28
+Test4/Talking_head/Test04_Talking_Head_D08.avi, 13.913037, 14.950007176003444, 38.31
+Test6/City/Test06_City_D16.avi, 14.998655, 29.736, 58.33
+Test3/Talking_head/Test03_Talking_Head_D02.avi, 15.081748, 29.97000002997, 36.32
+Test6/Talking_head/Test06_Talking_head_D15.avi, 14.993459, 29.813, 73.35
+Test5/City/Test05_City_D15.avi, 15.000167, 29.933, 71.29
+Test2/City/Test02_City_D07.avi, 15.816123, 29.526832509042592, 62.54
+Test1/City/Test01_City_D03.avi, 11.75584, 14.460897732531235, 17.96
+Test1/Traffic/Test01_Traffic_D04.avi, 21.209263999999997, 29.98689572656749, 25.72
+Test2/City/Test02_City_D03.avi, 16.039728, 14.96284725027756, 18.83
+Test6/Talking_head/Test06_Talking_head_D02.avi, 14.994377, 24.009, 57.48
+Test5/Talking_head/Test05_Talking_Head_D08.avi, 14.982325999999999, 24.896, 61.01
+Test4/Talking_head/Test04_Talking_Head_D03.avi, 14.24, 25.0, 77.09
+Test5/Talking_head/Test05_Talking_Head_D06.avi, 14.990884, 29.618, 71.0
+Test5/Talking_head/Test05_Talking_Head_D11.avi, 14.994187, 24.943, 56.29
+Test3/Newspaper/Test03_Newspaper_D04.avi, 15.017282, 23.572841258129685, 39.98
+Test5/Television/Test05_Television_D02.avi, 21.157850999999997, 24.01, 42.29
+Test5/Television/Test05_Television_D16.avi, 20.881045, 29.692, 45.88
+Test6/Television/Test06_Television_D15.avi, 21.997242, 29.731, 33.46
+Test5/Talking_head/Test05_Talking_Head_D14.avi, 14.985776, 24.957, 67.29
+Test3/Newspaper/Test03_Newspaper_D03.avi, 14.999984999999999, 30.00003000003, 55.89
+Test3/City/Test03_City_D05.avi, 15.015015, 29.97000002997, 22.47
+Test5/Talking_head/Test05_Talking_Head_D12.avi, 15.009606, 24.984, 68.36
+Test4/Newspaper/Test04_Newspaper_D02.avi, 15.0, 25.0, 82.05
+Test5/City/Test05_City_D16.avi, 15.016328999999999, 29.701, 76.85
+Test4/City/Test04_City_D09.avi, 15.011643, 30.109961579689024, 62.08
+Test3/City/Test03_City_D07.avi, 14.912008, 30.109961579689024, 35.74
+Test4/Talking_head/Test04_Talking_Head_D04.avi, 15.066652, 30.00003000003, 83.73
+Test5/City/Test05_City_D07.avi, 15.002127999999999, 28.196, 66.67
+Test4/City/Test04_City_D14.avi, 14.997898, 23.669983620371333, 42.63
+Test5/Television/Test05_Television_D08.avi, 21.296433, 20.379, 27.84
+Test6/Television/Test06_Television_D16.avi, 21.986746, 28.972, 20.54
+Test2/Talking_head/Test02_Talking_Head_D08.avi, 16.572551, 29.8686674691382, 35.7
+Test3/Talking_head/Test03_Talking_Head_D13.avi, 15.673661, 29.54000307216032, 32.72
+Test4/Talking_head/Test04_Talking_Head_D07.avi, 14.515042999999999, 14.950007176003444, 31.15
+Test3/Newspaper/Test03_Newspaper_D12.avi, 15.120292, 29.099967117037156, 41.32
+Test4/Newspaper/Test04_Newspaper_D03.avi, 15.0, 25.0, 81.33
+Test6/Talking_head/Test06_Talking_head_D06.avi, 14.980642999999999, 23.764, 61.99
+Test6/City/Test06_City_D15.avi, 15.006364, 29.854, 71.95
+Test4/Newspaper/Test04_Newspaper_D12.avi, 14.982835, 29.099967117037156, 37.95
+Test4/City/Test04_City_D11.avi, 14.985992, 14.280002284800366, -1.11
+Test5/Talking_head/Test05_Talking_Head_D09.avi, 15.0, 30.0, 68.07
+Test4/City/Test04_City_D06.avi, 14.990694, 10.739996498761142, -6.5
+Test6/City/Test06_City_D02.avi, 15.015015, 29.97, 72.11
+Test4/City/Test04_City_D04.avi, 14.999984999999999, 30.00003000003, 69.0
+Test6/Talking_head/Test06_Talking_head_D05.avi, 15.009051999999999, 17.123, 45.07
+Test6/Television/Test06_Television_D12.avi, 22.014969999999998, 24.983, 22.48
+Test3/City/Test03_City_D09.avi, 15.006895, 28.92003898421255, 30.86
+Test4/Talking_head/Test04_Talking_Head_D10.avi, 13.879999999999999, 25.0, 61.18
+Test6/Television/Test06_Television_D10.avi, 22.014595999999997, 24.938, 14.77
+Test4/Newspaper/Test04_Newspaper_D04.avi, 14.999984999999999, 30.00003000003, 82.52
+Test1/Traffic/Test01_Traffic_D02.avi, 22.044, 15.151515151515152, 29.57
+Test4/City/Test04_City_D07.avi, 15.020033, 14.979994217722233, 16.12
+Test3/Talking_head/Test03_Talking_Head_D07.avi, 16.058394, 30.140000301400004, 35.29
+Test4/Talking_head/Test04_Talking_Head_D12.avi, 15.945034999999999, 29.099967117037156, 17.06
+Test1/Talking_head/Test01_Talking_Head_D04.avi, 16.930918, 30.06334346468008, 30.87
+Test6/Talking_head/Test06_Talking_head_D16.avi, 15.007688, 29.918, 33.02
+Test1/Talking_head/Test01_Talking_Head_D07.avi, 16.832297, 29.52657096120799, 35.99
+Test1/Traffic/Test01_Traffic_D05.avi, 24.218208, 14.86484879475806, 30.21
+Test3/City/Test03_City_D06.avi, 15.100672999999999, 14.89999865900012, 12.85
+Test6/City/Test06_City_D04.avi, 15.0, 25.0, 67.94
+Test5/Talking_head/Test05_Talking_Head_D01.avi, 15.000487, 30.799, 46.8
+Test1/Traffic/Test01_Traffic_D09.avi, 24.799999999999997, 25.0, 60.7
+Test5/City/Test05_City_D02.avi, 15.015015, 29.97, 76.43
+Test5/City/Test05_City_D14.avi, 15.010318, 30.046, 64.27
+Test1/City/Test01_City_D02.avi, 11.418, 15.151515151515152, 24.86
+Test2/Talking_head/Test02_Talking_Head_D03.avi, 17.592412, 14.949627231044994, 22.11
+Test5/Television/Test05_Television_D10.avi, 21.294514, 24.936, 22.5
+Test4/Newspaper/Test04_Newspaper_D05.avi, 15.015015, 29.97000002997, 79.36
+Test4/City/Test04_City_D12.avi, 15.017199, 29.099967117037156, 35.17
+Test6/Television/Test06_Television_D04.avi, 22.0, 25.0, 39.1
+Test5/Television/Test05_Television_D14.avi, 20.645419999999998, 24.945, 18.45
+Test5/Television/Test05_Television_D04.avi, 21.020144, 23.977, 55.0
+Test6/Television/Test06_Television_D07.avi, 21.985433, 14.828, 22.59
+Test3/Newspaper/Test03_Newspaper_D02.avi, 15.015015, 29.97000002997, 36.79
+Test6/Talking_head/Test06_Talking_head_D13.avi, 14.994187, 24.943, 47.76
+Test1/City/Test01_City_D08.avi, 11.799999999999999, 25.0, 63.68
+Test2/City/Test02_City_D02.avi, 15.923357, 29.95599464386816, 9.99
+Test6/Television/Test06_Television_D01.avi, 22.013702, 30.799, 10.93
+Test2/Traffic/Test02_Traffic_D04.avi, 24.58421, 14.846928170561512, 24.69
+Test5/Talking_head/Test05_Talking_Head_D10.avi, 14.997193, 24.938, 60.05
+Test4/City/Test04_City_D13.avi, 15.015015, 29.97000002997, 52.28
+Test3/Newspaper/Test03_Newspaper_D01.avi, 14.999849999999999, 30.00030000300003, 50.49
+Test1/Traffic/Test01_Traffic_D06.avi, 21.899675, 28.173934602663, 43.36
+Test6/Talking_head/Test06_Talking_head_D14.avi, 14.985049, 29.763, 58.85
+Test2/City/Test02_City_D09.avi, 16.08, 25.0, 56.28
+Test6/City/Test06_City_D14.avi, 15.009381, 29.848, 72.46
+Test1/City/Test01_City_D09.avi, 12.32, 25.0, 64.95
+Test4/Talking_head/Test04_Talking_Head_D06.avi, 14.606734999999999, 9.790004405501982, 1.97
+Test6/City/Test06_City_D09.avi, 15.018661999999999, 14.914776964425274, -0.85
+Test6/City/Test06_City_D03.avi, 15.013511999999999, 29.973, 71.76
+Test2/Talking_head/Test02_Talking_Head_D04.avi, 17.359872, 14.746652509880256, 16.69
+Test3/City/Test03_City_D03.avi, 14.966652, 30.00003000003, 57.62
+Test6/Television/Test06_Television_D13.avi, 22.011066, 24.942, 23.7
+Test3/City/Test03_City_D01.avi, 14.788022, 31.579612202362156, 34.75
+Test3/Talking_head/Test03_Talking_Head_D05.avi, 16.116115999999998, 29.97000002997, 36.98
+Test2/Traffic/Test02_Traffic_D10.avi, 28.059312, 29.5801978323631, 27.35
+Test5/City/Test05_City_D13.avi, 14.9988, 25.002, 62.65
+Test5/City/Test05_City_D06.avi, 14.983801, 29.632, 75.32
+Test1/City/Test01_City_D07.avi, 12.572566, 28.474695961433873, 51.67
+Test6/Television/Test06_Television_D14.avi, 21.998995, 29.865, 31.73
+Test5/Talking_head/Test05_Talking_Head_D13.avi, 14.9988, 25.002, 61.84
+Test3/Talking_head/Test03_Talking_Head_D04.avi, 15.893785999999999, 14.09355867994092, 27.86
+Test5/Television/Test05_Television_D09.avi, 20.666667, 30.0, 33.59
+Test2/Traffic/Test02_Traffic_D09.avi, 27.04, 25.0, 64.3

examplar_data_labels/DIVIDE_MaxWell/val_labels.txt

@@ -0,0 +1,909 @@
+0000.mp4, 1.114285714, 1.057142857, 1.028571429
+0012.mp4, 1.538461538, 1.423076923, 1.384615385
+0014.mp4, 1.423076923, 1.653846154, 1.423076923
+0015.mp4, 1.576923077, 1.5, 1.423076923
+0016.mp4, 1.448275862, 1.793103448, 1.448275862
+0020.mp4, 1.423076923, 1.576923077, 1.461538462
+0022.mp4, 1.551724138, 1.724137931, 1.482758621
+0028.mp4, 1.538461538, 1.615384615, 1.5
+0030.mp4, 1.357142857, 1.785714286, 1.5
+0036.mp4, 1.428571429, 1.928571429, 1.535714286
+0039.mp4, 1.642857143, 1.678571429, 1.571428571
+0040.mp4, 1.657142857, 1.914285714, 1.571428571
+0042.mp4, 1.5, 1.821428571, 1.571428571
+0043.mp4, 1.769230769, 1.538461538, 1.576923077
+0047.mp4, 1.428571429, 1.857142857, 1.6
+0048.mp4, 1.6, 1.971428571, 1.6
+0049.mp4, 1.538461538, 1.615384615, 1.615384615
+0052.mp4, 1.514285714, 2.2, 1.628571429
+0054.mp4, 1.607142857, 1.892857143, 1.642857143
+0061.mp4, 1.586206897, 1.896551724, 1.689655172
+0071.mp4, 1.742857143, 1.942857143, 1.742857143
+0073.mp4, 1.714285714, 1.964285714, 1.75
+0078.mp4, 1.615384615, 2.153846154, 1.769230769
+0081.mp4, 1.485714286, 2.314285714, 1.771428571
+0082.mp4, 1.607142857, 2.178571429, 1.785714286
+0097.mp4, 2.192307692, 1.692307692, 1.807692308
+0098.mp4, 1.692307692, 2.076923077, 1.807692308
+0100.mp4, 1.714285714, 2.142857143, 1.821428571
+0101.mp4, 1.823529412, 2.088235294, 1.823529412
+0103.mp4, 1.5, 2.384615385, 1.846153846
+0108.mp4, 1.821428571, 2.25, 1.857142857
+0112.mp4, 1.75862069, 2.275862069, 1.862068966
+0119.mp4, 1.692307692, 1.884615385, 1.884615385
+0122.mp4, 1.730769231, 2.307692308, 1.884615385
+0131.mp4, 1.655172414, 2.068965517, 1.896551724
+0138.mp4, 1.828571429, 2.314285714, 1.914285714
+0144.mp4, 1.642857143, 2.535714286, 1.928571429
+0145.mp4, 1.75862069, 2.034482759, 1.931034483
+0158.mp4, 1.769230769, 2.423076923, 1.961538462
+0162.mp4, 2.0, 2.230769231, 1.961538462
+0166.mp4, 1.964285714, 2.75, 1.964285714
+0170.mp4, 1.941176471, 2.235294118, 1.970588235
+0176.mp4, 2.071428571, 2.142857143, 2.0
+0180.mp4, 1.857142857, 2.342857143, 2.0
+0182.mp4, 1.896551724, 2.482758621, 2.0
+0184.mp4, 1.571428571, 2.571428571, 2.0
+0192.mp4, 1.678571429, 2.821428571, 2.0
+0196.mp4, 1.911764706, 2.235294118, 2.0
+0198.mp4, 1.794117647, 2.352941176, 2.0
+0202.mp4, 2.034482759, 2.413793103, 2.034482759
+0212.mp4, 1.923076923, 2.115384615, 2.038461538
+0220.mp4, 2.085714286, 2.571428571, 2.057142857
+0221.mp4, 1.971428571, 2.571428571, 2.057142857
+0225.mp4, 2.117647059, 2.235294118, 2.058823529
+0227.mp4, 1.970588235, 2.411764706, 2.058823529
+0228.mp4, 1.882352941, 2.529411765, 2.058823529
+0232.mp4, 1.75, 2.571428571, 2.071428571
+0236.mp4, 2.0, 2.214285714, 2.071428571
+0243.mp4, 2.057142857, 2.257142857, 2.085714286
+0247.mp4, 1.942857143, 2.6, 2.085714286
+0250.mp4, 1.794117647, 2.676470588, 2.088235294
+0251.mp4, 2.0, 2.235294118, 2.088235294
+0252.mp4, 2.058823529, 2.588235294, 2.088235294
+0256.mp4, 2.103448276, 2.344827586, 2.103448276
+0261.mp4, 1.821428571, 2.607142857, 2.107142857
+0264.mp4, 2.0, 2.321428571, 2.107142857
+0266.mp4, 1.914285714, 2.457142857, 2.114285714
+0268.mp4, 1.942857143, 2.714285714, 2.114285714
+0288.mp4, 1.928571429, 2.392857143, 2.142857143
+0294.mp4, 2.029411765, 2.588235294, 2.147058824
+0296.mp4, 1.970588235, 2.382352941, 2.147058824
+0297.mp4, 2.235294118, 2.264705882, 2.147058824
+0300.mp4, 2.088235294, 2.411764706, 2.147058824
+0310.mp4, 2.068965517, 2.620689655, 2.172413793
+0319.mp4, 1.928571429, 2.785714286, 2.178571429
+0321.mp4, 1.892857143, 3.0, 2.178571429
+0331.mp4, 2.107142857, 2.464285714, 2.178571429
+0337.mp4, 2.230769231, 2.615384615, 2.192307692
+0352.mp4, 2.088235294, 2.441176471, 2.205882353
+0356.mp4, 2.068965517, 2.620689655, 2.206896552
+0362.mp4, 2.24137931, 2.655172414, 2.206896552
+0364.mp4, 2.142857143, 2.571428571, 2.214285714
+0371.mp4, 2.028571429, 2.657142857, 2.228571429
+0372.mp4, 2.0, 2.828571429, 2.228571429
+0379.mp4, 2.423076923, 2.153846154, 2.230769231
+0384.mp4, 2.176470588, 2.588235294, 2.235294118
+0388.mp4, 2.275862069, 2.344827586, 2.24137931
+0394.mp4, 2.24137931, 2.379310345, 2.24137931
+0402.mp4, 2.142857143, 2.5, 2.25
+0407.mp4, 2.071428571, 2.857142857, 2.25
+0412.mp4, 2.142857143, 2.771428571, 2.257142857
+0413.mp4, 2.142857143, 2.857142857, 2.257142857
+0415.mp4, 2.205882353, 2.470588235, 2.264705882
+0419.mp4, 2.153846154, 2.730769231, 2.269230769
+0425.mp4, 2.153846154, 2.461538462, 2.269230769
+0426.mp4, 2.153846154, 2.615384615, 2.269230769
+0430.mp4, 2.153846154, 2.807692308, 2.269230769
+0431.mp4, 2.0, 2.807692308, 2.269230769
+0435.mp4, 2.423076923, 2.192307692, 2.269230769
+0438.mp4, 2.153846154, 2.692307692, 2.269230769
+0441.mp4, 1.896551724, 2.827586207, 2.275862069
+0445.mp4, 2.068965517, 2.965517241, 2.275862069
+0449.mp4, 2.214285714, 2.678571429, 2.285714286
+0457.mp4, 2.142857143, 2.542857143, 2.285714286
+0459.mp4, 2.107142857, 2.571428571, 2.285714286
+0463.mp4, 2.214285714, 2.785714286, 2.285714286
+0472.mp4, 2.346153846, 2.461538462, 2.307692308
+0476.mp4, 2.115384615, 2.692307692, 2.307692308
+0478.mp4, 2.076923077, 2.884615385, 2.307692308
+0484.mp4, 2.615384615, 2.538461538, 2.307692308
+0487.mp4, 2.038461538, 2.730769231, 2.307692308
+0507.mp4, 2.214285714, 2.642857143, 2.321428571
+0514.mp4, 2.205882353, 2.411764706, 2.323529412
+0526.mp4, 2.103448276, 2.75862069, 2.344827586
+0533.mp4, 2.346153846, 2.576923077, 2.346153846
+0540.mp4, 2.461538462, 2.5, 2.346153846
+0544.mp4, 1.923076923, 2.730769231, 2.346153846
+0545.mp4, 2.153846154, 2.730769231, 2.346153846
+0551.mp4, 2.307692308, 2.884615385, 2.346153846
+0553.mp4, 2.153846154, 2.923076923, 2.346153846
+0554.mp4, 2.269230769, 2.461538462, 2.346153846
+0557.mp4, 2.264705882, 2.588235294, 2.352941176
+0558.mp4, 2.357142857, 2.392857143, 2.357142857
+0571.mp4, 2.107142857, 2.857142857, 2.357142857
+0578.mp4, 2.142857143, 2.685714286, 2.371428571
+0583.mp4, 2.344827586, 2.827586207, 2.379310345
+0585.mp4, 2.206896552, 2.931034483, 2.379310345
+0589.mp4, 1.931034483, 2.75862069, 2.379310345
+0593.mp4, 2.264705882, 2.852941176, 2.382352941
+0595.mp4, 2.382352941, 2.558823529, 2.382352941
+0601.mp4, 2.692307692, 2.269230769, 2.384615385
+0602.mp4, 2.192307692, 2.538461538, 2.384615385
+0606.mp4, 1.923076923, 2.5, 2.384615385
+0619.mp4, 2.0, 2.785714286, 2.392857143
+0627.mp4, 2.142857143, 3.0, 2.392857143
+0629.mp4, 2.228571429, 2.885714286, 2.4
+0635.mp4, 2.352941176, 2.558823529, 2.411764706
+0639.mp4, 2.323529412, 2.647058824, 2.411764706
+0642.mp4, 2.352941176, 2.794117647, 2.411764706
+0649.mp4, 2.24137931, 2.655172414, 2.413793103
+0657.mp4, 2.115384615, 2.538461538, 2.423076923
+0666.mp4, 2.153846154, 2.538461538, 2.423076923
+0669.mp4, 2.269230769, 2.692307692, 2.423076923
+0672.mp4, 2.307692308, 2.730769231, 2.423076923
+0683.mp4, 2.807692308, 2.230769231, 2.423076923
+0684.mp4, 2.538461538, 2.615384615, 2.423076923
+0686.mp4, 2.321428571, 2.642857143, 2.428571429
+0689.mp4, 2.357142857, 2.821428571, 2.428571429
+0692.mp4, 2.357142857, 2.821428571, 2.428571429
+0695.mp4, 2.357142857, 2.428571429, 2.428571429
+0699.mp4, 2.25, 2.678571429, 2.428571429
+0700.mp4, 2.114285714, 2.742857143, 2.428571429
+0704.mp4, 2.171428571, 2.828571429, 2.428571429
+0715.mp4, 2.24137931, 2.75862069, 2.448275862
+0716.mp4, 2.655172414, 2.275862069, 2.448275862
+0722.mp4, 2.034482759, 2.896551724, 2.448275862
+0723.mp4, 2.413793103, 2.931034483, 2.448275862
+0726.mp4, 2.371428571, 2.714285714, 2.457142857
+0729.mp4, 1.914285714, 3.171428571, 2.457142857
+0730.mp4, 2.346153846, 2.5, 2.461538462
+0731.mp4, 2.423076923, 2.538461538, 2.461538462
+0732.mp4, 2.307692308, 2.538461538, 2.461538462
+0741.mp4, 2.423076923, 2.807692308, 2.461538462
+0742.mp4, 2.192307692, 2.884615385, 2.461538462
+0750.mp4, 2.5, 2.576923077, 2.461538462
+0753.mp4, 2.730769231, 2.346153846, 2.461538462
+0754.mp4, 2.307692308, 2.576923077, 2.461538462
+0757.mp4, 2.461538462, 2.769230769, 2.461538462
+0760.mp4, 2.230769231, 2.923076923, 2.461538462
+0774.mp4, 2.25, 2.642857143, 2.464285714
+0775.mp4, 2.321428571, 2.857142857, 2.464285714
+0790.mp4, 2.2, 2.8, 2.485714286
+0791.mp4, 2.257142857, 2.857142857, 2.485714286
+0793.mp4, 2.371428571, 2.885714286, 2.485714286
+0798.mp4, 2.5, 2.653846154, 2.5
+0799.mp4, 2.5, 2.653846154, 2.5
+0800.mp4, 2.192307692, 2.653846154, 2.5
+0805.mp4, 2.384615385, 2.730769231, 2.5
+0807.mp4, 2.214285714, 2.75, 2.5
+0809.mp4, 2.269230769, 2.769230769, 2.5
+0813.mp4, 2.269230769, 2.884615385, 2.5
+0821.mp4, 2.192307692, 3.115384615, 2.5
+0824.mp4, 2.653846154, 2.461538462, 2.5
+0826.mp4, 2.346153846, 3.0, 2.5
+0828.mp4, 2.384615385, 2.576923077, 2.5
+0834.mp4, 2.178571429, 2.892857143, 2.5
+0838.mp4, 2.178571429, 3.071428571, 2.5
+0841.mp4, 2.441176471, 2.882352941, 2.5
+0843.mp4, 2.558823529, 2.735294118, 2.5
+0854.mp4, 2.657142857, 2.485714286, 2.514285714
+0857.mp4, 2.285714286, 2.714285714, 2.514285714
+0859.mp4, 2.628571429, 2.885714286, 2.514285714
+0861.mp4, 2.228571429, 3.0, 2.514285714
+0869.mp4, 2.172413793, 2.827586207, 2.517241379
+0872.mp4, 2.294117647, 2.735294118, 2.529411765
+0887.mp4, 2.535714286, 2.607142857, 2.535714286
+0891.mp4, 2.357142857, 2.785714286, 2.535714286
+0893.mp4, 2.392857143, 2.821428571, 2.535714286
+0894.mp4, 2.214285714, 2.821428571, 2.535714286
+0900.mp4, 2.615384615, 2.692307692, 2.538461538
+0905.mp4, 2.846153846, 2.423076923, 2.538461538
+0908.mp4, 2.461538462, 2.615384615, 2.538461538
+0911.mp4, 2.384615385, 2.730769231, 2.538461538
+0915.mp4, 2.461538462, 2.846153846, 2.538461538
+0917.mp4, 2.230769231, 2.846153846, 2.538461538
+0929.mp4, 2.457142857, 2.828571429, 2.542857143
+0930.mp4, 2.314285714, 2.914285714, 2.542857143
+0939.mp4, 2.275862069, 2.965517241, 2.551724138
+0949.mp4, 2.529411765, 2.558823529, 2.558823529
+0950.mp4, 2.588235294, 2.647058824, 2.558823529
+0952.mp4, 2.382352941, 2.617647059, 2.558823529
+0955.mp4, 2.647058824, 2.529411765, 2.558823529
+0959.mp4, 2.607142857, 2.607142857, 2.571428571
+0962.mp4, 2.571428571, 2.785714286, 2.571428571
+0963.mp4, 2.371428571, 2.8, 2.571428571
+0968.mp4, 2.285714286, 2.857142857, 2.571428571
+0970.mp4, 2.485714286, 2.885714286, 2.571428571
+0983.mp4, 2.285714286, 2.785714286, 2.571428571
+0995.mp4, 2.384615385, 2.807692308, 2.576923077
+1006.mp4, 2.346153846, 2.961538462, 2.576923077
+1010.mp4, 2.653846154, 2.384615385, 2.576923077
+1011.mp4, 3.115384615, 2.384615385, 2.576923077
+1017.mp4, 2.75862069, 2.517241379, 2.586206897
+1019.mp4, 2.517241379, 2.793103448, 2.586206897
+1021.mp4, 2.379310345, 2.827586207, 2.586206897
+1027.mp4, 2.379310345, 3.0, 2.586206897
+1039.mp4, 2.5, 2.852941176, 2.588235294
+1041.mp4, 2.457142857, 2.685714286, 2.6
+1043.mp4, 2.257142857, 2.8, 2.6
+1047.mp4, 2.142857143, 3.028571429, 2.6
+1057.mp4, 2.464285714, 2.535714286, 2.607142857
+1060.mp4, 2.357142857, 2.892857143, 2.607142857
+1064.mp4, 2.615384615, 2.807692308, 2.615384615
+1068.mp4, 2.461538462, 3.076923077, 2.615384615
+1069.mp4, 2.538461538, 2.615384615, 2.615384615
+1070.mp4, 2.615384615, 2.730769231, 2.615384615
+1075.mp4, 2.461538462, 2.846153846, 2.615384615
+1079.mp4, 2.384615385, 2.961538462, 2.615384615
+1091.mp4, 2.384615385, 3.153846154, 2.615384615
+1098.mp4, 2.769230769, 2.615384615, 2.615384615
+1102.mp4, 2.5, 2.794117647, 2.617647059
+1111.mp4, 2.441176471, 2.852941176, 2.617647059
+1122.mp4, 2.586206897, 2.965517241, 2.620689655
+1128.mp4, 2.514285714, 2.885714286, 2.628571429
+1134.mp4, 2.428571429, 2.892857143, 2.642857143
+1149.mp4, 2.392857143, 2.964285714, 2.642857143
+1156.mp4, 2.5, 2.970588235, 2.647058824
+1161.mp4, 2.615384615, 2.769230769, 2.653846154
+1168.mp4, 2.346153846, 2.884615385, 2.653846154
+1174.mp4, 2.384615385, 3.0, 2.653846154
+1181.mp4, 2.384615385, 3.269230769, 2.653846154
+1187.mp4, 2.346153846, 3.076923077, 2.653846154
+1190.mp4, 3.153846154, 2.230769231, 2.653846154
+1198.mp4, 2.538461538, 3.038461538, 2.653846154
+1200.mp4, 2.384615385, 3.076923077, 2.653846154
+1204.mp4, 2.413793103, 3.0, 2.655172414
+1214.mp4, 2.551724138, 2.862068966, 2.655172414
+1215.mp4, 2.714285714, 2.828571429, 2.657142857
+1229.mp4, 2.571428571, 3.071428571, 2.678571429
+1231.mp4, 2.571428571, 2.964285714, 2.678571429
+1232.mp4, 2.392857143, 2.857142857, 2.678571429
+1236.mp4, 2.5, 2.857142857, 2.678571429
+1242.mp4, 2.571428571, 3.214285714, 2.678571429
+1243.mp4, 2.607142857, 2.714285714, 2.678571429
+1247.mp4, 2.571428571, 2.928571429, 2.678571429
+1248.mp4, 2.535714286, 2.964285714, 2.678571429
+1249.mp4, 2.428571429, 3.0, 2.678571429
+1256.mp4, 2.448275862, 2.931034483, 2.689655172
+1264.mp4, 2.448275862, 3.206896552, 2.689655172
+1278.mp4, 2.538461538, 2.807692308, 2.692307692
+1281.mp4, 2.538461538, 2.846153846, 2.692307692
+1283.mp4, 2.538461538, 3.0, 2.692307692
+1286.mp4, 2.423076923, 3.076923077, 2.692307692
+1290.mp4, 2.5, 3.192307692, 2.692307692
+1293.mp4, 2.538461538, 2.923076923, 2.692307692
+1300.mp4, 2.923076923, 2.615384615, 2.692307692
+1301.mp4, 2.153846154, 3.269230769, 2.692307692
+1311.mp4, 2.558823529, 3.058823529, 2.705882353
+1312.mp4, 2.764705882, 2.882352941, 2.705882353
+1329.mp4, 2.75, 3.035714286, 2.714285714
+1334.mp4, 2.464285714, 3.0, 2.714285714
+1335.mp4, 2.607142857, 2.892857143, 2.714285714
+1336.mp4, 2.321428571, 3.178571429, 2.714285714
+1339.mp4, 2.5, 3.035714286, 2.714285714
+1346.mp4, 2.392857143, 2.892857143, 2.714285714
+1349.mp4, 2.392857143, 3.142857143, 2.714285714
+1357.mp4, 2.827586207, 2.827586207, 2.724137931
+1369.mp4, 2.576923077, 2.807692308, 2.730769231
+1371.mp4, 2.615384615, 2.846153846, 2.730769231
+1384.mp4, 2.538461538, 3.0, 2.730769231
+1389.mp4, 2.5, 3.076923077, 2.730769231
+1403.mp4, 2.115384615, 2.961538462, 2.730769231
+1404.mp4, 2.423076923, 3.0, 2.730769231
+1407.mp4, 3.038461538, 2.5, 2.730769231
+1409.mp4, 2.615384615, 2.846153846, 2.730769231
+1410.mp4, 2.807692308, 2.923076923, 2.730769231
+1416.mp4, 2.461538462, 3.115384615, 2.730769231
+1419.mp4, 2.615384615, 3.307692308, 2.730769231
+1420.mp4, 2.735294118, 3.058823529, 2.735294118
+1421.mp4, 2.588235294, 2.941176471, 2.735294118
+1430.mp4, 2.4, 2.885714286, 2.742857143
+1435.mp4, 2.75, 2.571428571, 2.75
+1446.mp4, 2.464285714, 3.178571429, 2.75
+1447.mp4, 2.5, 3.178571429, 2.75
+1452.mp4, 2.571428571, 3.035714286, 2.75
+1454.mp4, 2.321428571, 3.071428571, 2.75
+1456.mp4, 2.321428571, 3.107142857, 2.75
+1460.mp4, 2.724137931, 3.0, 2.75862069
+1461.mp4, 2.724137931, 3.034482759, 2.75862069
+1465.mp4, 2.517241379, 3.0, 2.75862069
+1466.mp4, 2.586206897, 3.034482759, 2.75862069
+1471.mp4, 2.655172414, 2.862068966, 2.75862069
+1473.mp4, 2.75862069, 3.068965517, 2.75862069
+1487.mp4, 2.615384615, 3.192307692, 2.769230769
+1491.mp4, 2.730769231, 2.846153846, 2.769230769
+1495.mp4, 2.5, 2.923076923, 2.769230769
+1501.mp4, 2.730769231, 3.0, 2.769230769
+1502.mp4, 2.5, 3.076923077, 2.769230769
+1512.mp4, 2.5, 3.5, 2.769230769
+1524.mp4, 2.514285714, 3.371428571, 2.771428571
+1530.mp4, 2.607142857, 2.928571429, 2.785714286
+1532.mp4, 2.642857143, 3.0, 2.785714286
+1538.mp4, 2.785714286, 2.892857143, 2.785714286
+1541.mp4, 2.642857143, 2.964285714, 2.785714286
+1546.mp4, 2.357142857, 3.142857143, 2.785714286
+1570.mp4, 2.551724138, 3.172413793, 2.793103448
+1572.mp4, 2.75862069, 3.206896552, 2.793103448
+1593.mp4, 2.653846154, 2.769230769, 2.807692308
+1594.mp4, 2.615384615, 2.846153846, 2.807692308
+1596.mp4, 2.769230769, 2.846153846, 2.807692308
+1598.mp4, 2.807692308, 2.884615385, 2.807692308
+1600.mp4, 2.807692308, 2.923076923, 2.807692308
+1602.mp4, 2.807692308, 3.0, 2.807692308
+1604.mp4, 2.576923077, 3.0, 2.807692308
+1606.mp4, 2.615384615, 3.038461538, 2.807692308
+1607.mp4, 2.846153846, 3.038461538, 2.807692308
+1611.mp4, 2.730769231, 3.038461538, 2.807692308
+1615.mp4, 2.615384615, 3.076923077, 2.807692308
+1620.mp4, 2.576923077, 3.153846154, 2.807692308
+1622.mp4, 2.576923077, 3.153846154, 2.807692308
+1625.mp4, 2.846153846, 2.730769231, 2.807692308
+1632.mp4, 3.038461538, 2.423076923, 2.807692308
+1635.mp4, 2.423076923, 3.192307692, 2.807692308
+1637.mp4, 2.730769231, 2.961538462, 2.807692308
+1639.mp4, 2.884615385, 2.692307692, 2.807692308
+1643.mp4, 3.0, 2.923076923, 2.807692308
+1649.mp4, 2.846153846, 3.038461538, 2.807692308
+1653.mp4, 2.692307692, 3.153846154, 2.807692308
+1654.mp4, 2.730769231, 3.230769231, 2.807692308
+1661.mp4, 2.607142857, 2.964285714, 2.821428571
+1662.mp4, 2.678571429, 3.0, 2.821428571
+1665.mp4, 2.464285714, 3.178571429, 2.821428571
+1668.mp4, 2.714285714, 3.0, 2.821428571
+1678.mp4, 2.678571429, 2.678571429, 2.821428571
+1683.mp4, 2.647058824, 3.058823529, 2.823529412
+1684.mp4, 2.588235294, 3.029411765, 2.823529412
+1685.mp4, 2.558823529, 3.147058824, 2.823529412
+1688.mp4, 2.794117647, 2.911764706, 2.823529412
+1695.mp4, 2.655172414, 3.137931034, 2.827586207
+1705.mp4, 2.827586207, 3.068965517, 2.827586207
+1707.mp4, 2.542857143, 2.857142857, 2.828571429
+1722.mp4, 2.653846154, 3.269230769, 2.846153846
+1727.mp4, 2.769230769, 2.884615385, 2.846153846
+1729.mp4, 2.846153846, 2.923076923, 2.846153846
+1730.mp4, 2.846153846, 2.923076923, 2.846153846
+1731.mp4, 2.730769231, 2.923076923, 2.846153846
+1739.mp4, 2.923076923, 3.038461538, 2.846153846
+1752.mp4, 2.769230769, 3.423076923, 2.846153846
+1759.mp4, 2.730769231, 3.115384615, 2.846153846
+1766.mp4, 2.807692308, 2.846153846, 2.846153846
+1767.mp4, 2.653846154, 3.307692308, 2.846153846
+1773.mp4, 2.764705882, 3.058823529, 2.852941176
+1778.mp4, 2.794117647, 2.970588235, 2.852941176
+1795.mp4, 2.685714286, 3.2, 2.857142857
+1797.mp4, 2.642857143, 3.25, 2.857142857
+1807.mp4, 2.857142857, 2.714285714, 2.857142857
+1808.mp4, 2.678571429, 3.0, 2.857142857
+1820.mp4, 2.75862069, 3.172413793, 2.862068966
+1824.mp4, 2.724137931, 2.931034483, 2.862068966
+1848.mp4, 2.576923077, 2.923076923, 2.884615385
+1852.mp4, 2.730769231, 3.0, 2.884615385
+1853.mp4, 2.846153846, 3.0, 2.884615385
+1855.mp4, 2.269230769, 3.076923077, 2.884615385
+1863.mp4, 2.730769231, 3.153846154, 2.884615385
+1868.mp4, 2.730769231, 3.076923077, 2.884615385
+1887.mp4, 2.846153846, 3.192307692, 2.884615385
+1893.mp4, 2.6, 3.457142857, 2.885714286
+1896.mp4, 2.464285714, 3.214285714, 2.892857143
+1897.mp4, 2.607142857, 3.035714286, 2.892857143
+1909.mp4, 2.678571429, 2.857142857, 2.892857143
+1911.mp4, 2.892857143, 2.928571429, 2.892857143
+1913.mp4, 2.678571429, 3.25, 2.892857143
+1915.mp4, 2.551724138, 2.896551724, 2.896551724
+1928.mp4, 2.586206897, 3.137931034, 2.896551724
+1929.mp4, 2.862068966, 3.172413793, 2.896551724
+1941.mp4, 2.857142857, 2.971428571, 2.914285714
+1951.mp4, 2.692307692, 3.115384615, 2.923076923
+1952.mp4, 2.615384615, 3.153846154, 2.923076923
+1958.mp4, 2.769230769, 3.269230769, 2.923076923
+1963.mp4, 2.884615385, 2.807692308, 2.923076923
+1969.mp4, 2.576923077, 3.0, 2.923076923
+1972.mp4, 2.846153846, 3.076923077, 2.923076923
+1974.mp4, 2.653846154, 3.115384615, 2.923076923
+1981.mp4, 2.653846154, 3.153846154, 2.923076923
+1986.mp4, 2.884615385, 3.192307692, 2.923076923
+1987.mp4, 2.538461538, 3.192307692, 2.923076923
+1992.mp4, 2.769230769, 2.846153846, 2.923076923
+1996.mp4, 2.692307692, 3.269230769, 2.923076923
+2000.mp4, 2.807692308, 3.192307692, 2.923076923
+2013.mp4, 2.678571429, 3.071428571, 2.928571429
+2026.mp4, 3.0, 3.0, 2.931034483
+2033.mp4, 2.655172414, 3.275862069, 2.931034483
+2043.mp4, 2.735294118, 3.147058824, 2.941176471
+2060.mp4, 2.714285714, 3.2, 2.942857143
+2061.mp4, 2.714285714, 3.371428571, 2.942857143
+2062.mp4, 3.153846154, 2.923076923, 2.961538462
+2068.mp4, 2.730769231, 3.076923077, 2.961538462
+2071.mp4, 2.923076923, 3.076923077, 2.961538462
+2073.mp4, 2.961538462, 3.153846154, 2.961538462
+2074.mp4, 2.884615385, 3.153846154, 2.961538462
+2080.mp4, 2.653846154, 3.269230769, 2.961538462
+2092.mp4, 2.769230769, 3.307692308, 2.961538462
+2093.mp4, 2.576923077, 3.269230769, 2.961538462
+2094.mp4, 2.769230769, 3.230769231, 2.961538462
+2096.mp4, 2.615384615, 3.076923077, 2.961538462
+2097.mp4, 2.653846154, 2.884615385, 2.961538462
+2104.mp4, 2.769230769, 3.115384615, 2.961538462
+2106.mp4, 2.884615385, 3.153846154, 2.961538462
+2112.mp4, 2.607142857, 3.0, 2.964285714
+2132.mp4, 2.607142857, 3.142857143, 2.964285714
+2134.mp4, 3.035714286, 2.75, 2.964285714
+2138.mp4, 2.714285714, 3.214285714, 2.964285714
+2149.mp4, 2.75862069, 3.137931034, 2.965517241
+2164.mp4, 2.75862069, 3.275862069, 2.965517241
+2178.mp4, 2.705882353, 2.941176471, 2.970588235
+2179.mp4, 2.882352941, 2.941176471, 2.970588235
+2188.mp4, 2.628571429, 3.171428571, 2.971428571
+2190.mp4, 2.8, 3.257142857, 2.971428571
+2191.mp4, 2.571428571, 3.257142857, 2.971428571
+2199.mp4, 2.771428571, 3.571428571, 2.971428571
+2203.mp4, 2.961538462, 3.0, 3.0
+2205.mp4, 2.923076923, 3.038461538, 3.0
+2208.mp4, 2.892857143, 3.071428571, 3.0
+2213.mp4, 3.115384615, 3.076923077, 3.0
+2215.mp4, 2.821428571, 3.107142857, 3.0
+2232.mp4, 2.807692308, 3.192307692, 3.0
+2239.mp4, 2.75, 3.214285714, 3.0
+2240.mp4, 2.653846154, 3.230769231, 3.0
+2255.mp4, 2.692307692, 3.269230769, 3.0
+2264.mp4, 2.846153846, 3.346153846, 3.0
+2265.mp4, 2.730769231, 3.346153846, 3.0
+2268.mp4, 2.807692308, 3.384615385, 3.0
+2284.mp4, 2.653846154, 3.192307692, 3.0
+2291.mp4, 2.615384615, 3.346153846, 3.0
+2303.mp4, 3.0, 3.107142857, 3.0
+2306.mp4, 2.714285714, 3.357142857, 3.0
+2315.mp4, 2.827586207, 3.172413793, 3.0
+2317.mp4, 2.551724138, 3.275862069, 3.0
+2320.mp4, 2.75862069, 3.137931034, 3.0
+2326.mp4, 2.970588235, 2.911764706, 3.0
+2327.mp4, 2.558823529, 3.205882353, 3.0
+2328.mp4, 3.0, 3.117647059, 3.0
+2330.mp4, 3.029411765, 3.147058824, 3.0
+2332.mp4, 2.911764706, 3.117647059, 3.0
+2333.mp4, 3.0, 3.088235294, 3.0
+2335.mp4, 2.882352941, 3.176470588, 3.0
+2340.mp4, 2.794117647, 3.117647059, 3.0
+2343.mp4, 2.8, 3.285714286, 3.028571429
+2347.mp4, 2.764705882, 3.352941176, 3.029411765
+2359.mp4, 3.0, 2.911764706, 3.029411765
+2371.mp4, 2.793103448, 3.24137931, 3.034482759
+2376.mp4, 2.793103448, 3.034482759, 3.034482759
+2379.mp4, 2.793103448, 3.137931034, 3.034482759
+2384.mp4, 2.821428571, 3.714285714, 3.035714286
+2385.mp4, 2.821428571, 3.357142857, 3.035714286
+2389.mp4, 2.928571429, 3.357142857, 3.035714286
+2390.mp4, 3.142857143, 3.214285714, 3.035714286
+2391.mp4, 2.714285714, 3.464285714, 3.035714286
+2395.mp4, 3.142857143, 3.035714286, 3.035714286
+2396.mp4, 2.892857143, 3.071428571, 3.035714286
+2399.mp4, 2.714285714, 3.178571429, 3.035714286
+2408.mp4, 2.884615385, 3.192307692, 3.038461538
+2420.mp4, 2.884615385, 3.038461538, 3.038461538
+2428.mp4, 2.730769231, 3.192307692, 3.038461538
+2439.mp4, 2.884615385, 3.384615385, 3.038461538
+2440.mp4, 3.0, 3.384615385, 3.038461538
+2456.mp4, 2.653846154, 3.346153846, 3.038461538
+2462.mp4, 3.115384615, 3.230769231, 3.038461538
+2463.mp4, 2.961538462, 3.115384615, 3.038461538
+2466.mp4, 2.628571429, 3.171428571, 3.057142857
+2471.mp4, 2.628571429, 3.4, 3.057142857
+2473.mp4, 2.714285714, 3.714285714, 3.057142857
+2482.mp4, 2.794117647, 3.264705882, 3.058823529
+2484.mp4, 2.735294118, 3.264705882, 3.058823529
+2503.mp4, 2.965517241, 3.310344828, 3.068965517
+2505.mp4, 3.137931034, 3.068965517, 3.068965517
+2508.mp4, 2.586206897, 3.24137931, 3.068965517
+2510.mp4, 3.034482759, 3.310344828, 3.068965517
+2512.mp4, 2.928571429, 3.107142857, 3.071428571
+2514.mp4, 2.75, 3.535714286, 3.071428571
+2516.mp4, 2.75, 3.285714286, 3.071428571
+2524.mp4, 2.785714286, 3.428571429, 3.071428571
+2527.mp4, 2.892857143, 3.285714286, 3.071428571
+2532.mp4, 3.115384615, 3.115384615, 3.076923077
+2535.mp4, 2.884615385, 3.153846154, 3.076923077
+2545.mp4, 3.0, 3.346153846, 3.076923077
+2554.mp4, 2.961538462, 3.153846154, 3.076923077
+2564.mp4, 3.038461538, 3.192307692, 3.076923077
+2568.mp4, 3.076923077, 3.230769231, 3.076923077
+2576.mp4, 3.057142857, 2.971428571, 3.085714286
+2579.mp4, 3.0, 3.257142857, 3.085714286
+2587.mp4, 2.714285714, 3.542857143, 3.085714286
+2589.mp4, 2.970588235, 3.323529412, 3.088235294
+2596.mp4, 3.147058824, 3.147058824, 3.088235294
+2597.mp4, 3.029411765, 3.264705882, 3.088235294
+2598.mp4, 3.0, 3.117647059, 3.088235294
+2599.mp4, 2.794117647, 3.205882353, 3.088235294
+2604.mp4, 3.103448276, 3.137931034, 3.103448276
+2610.mp4, 2.620689655, 3.379310345, 3.103448276
+2621.mp4, 3.310344828, 3.137931034, 3.103448276
+2623.mp4, 3.034482759, 3.034482759, 3.103448276
+2624.mp4, 2.965517241, 3.103448276, 3.103448276
+2627.mp4, 3.068965517, 2.931034483, 3.103448276
+2629.mp4, 2.931034483, 3.482758621, 3.103448276
+2642.mp4, 2.821428571, 3.357142857, 3.107142857
+2650.mp4, 2.964285714, 3.25, 3.107142857
+2657.mp4, 2.828571429, 3.085714286, 3.114285714
+2660.mp4, 3.142857143, 3.171428571, 3.114285714
+2663.mp4, 2.742857143, 3.371428571, 3.114285714
+2665.mp4, 2.771428571, 3.428571429, 3.114285714
+2673.mp4, 3.038461538, 3.269230769, 3.115384615
+2679.mp4, 2.961538462, 3.384615385, 3.115384615
+2686.mp4, 3.076923077, 3.076923077, 3.115384615
+2690.mp4, 2.884615385, 3.153846154, 3.115384615
+2699.mp4, 3.076923077, 3.230769231, 3.115384615
+2716.mp4, 2.923076923, 3.192307692, 3.115384615
+2720.mp4, 3.653846154, 2.653846154, 3.115384615
+2723.mp4, 2.923076923, 3.192307692, 3.115384615
+2727.mp4, 3.076923077, 2.730769231, 3.115384615
+2730.mp4, 2.923076923, 3.307692308, 3.115384615
+2737.mp4, 3.323529412, 3.176470588, 3.117647059
+2750.mp4, 2.896551724, 3.172413793, 3.137931034
+2751.mp4, 3.068965517, 3.137931034, 3.137931034
+2754.mp4, 3.0, 3.24137931, 3.137931034
+2756.mp4, 2.827586207, 3.137931034, 3.137931034
+2772.mp4, 3.035714286, 3.285714286, 3.142857143
+2783.mp4, 3.0, 3.25, 3.142857143
+2795.mp4, 3.0, 3.535714286, 3.142857143
+2797.mp4, 2.714285714, 3.628571429, 3.142857143
+2805.mp4, 3.117647059, 3.294117647, 3.147058824
+2816.mp4, 2.923076923, 3.346153846, 3.153846154
+2824.mp4, 2.923076923, 3.5, 3.153846154
+2834.mp4, 2.846153846, 3.076923077, 3.153846154
+2836.mp4, 3.0, 3.576923077, 3.153846154
+2838.mp4, 3.192307692, 2.961538462, 3.153846154
+2845.mp4, 3.307692308, 3.153846154, 3.153846154
+2854.mp4, 2.846153846, 3.423076923, 3.153846154
+2870.mp4, 2.971428571, 3.571428571, 3.171428571
+2875.mp4, 3.275862069, 3.068965517, 3.172413793
+2877.mp4, 3.034482759, 3.448275862, 3.172413793
+2887.mp4, 3.034482759, 3.344827586, 3.172413793
+2890.mp4, 3.029411765, 3.294117647, 3.176470588
+2891.mp4, 3.058823529, 3.205882353, 3.176470588
+2894.mp4, 3.088235294, 3.235294118, 3.176470588
+2901.mp4, 3.178571429, 3.142857143, 3.178571429
+2903.mp4, 2.821428571, 3.357142857, 3.178571429
+2905.mp4, 3.035714286, 3.25, 3.178571429
+2909.mp4, 2.964285714, 3.0, 3.178571429
+2911.mp4, 2.964285714, 3.464285714, 3.178571429
+2927.mp4, 3.115384615, 3.461538462, 3.192307692
+2933.mp4, 3.038461538, 3.576923077, 3.192307692
+2937.mp4, 3.038461538, 3.192307692, 3.192307692
+2939.mp4, 3.384615385, 3.230769231, 3.192307692
+2941.mp4, 2.923076923, 3.269230769, 3.192307692
+2944.mp4, 2.923076923, 3.346153846, 3.192307692
+2957.mp4, 3.076923077, 3.538461538, 3.192307692
+2970.mp4, 2.885714286, 3.314285714, 3.2
+2971.mp4, 2.914285714, 3.4, 3.2
+2977.mp4, 3.088235294, 3.441176471, 3.205882353
+2996.mp4, 2.896551724, 3.344827586, 3.206896552
+2997.mp4, 3.0, 3.448275862, 3.206896552
+2998.mp4, 2.75862069, 3.24137931, 3.206896552
+3003.mp4, 3.0, 3.357142857, 3.214285714
+3012.mp4, 3.25, 3.321428571, 3.214285714
+3017.mp4, 3.035714286, 3.357142857, 3.214285714
+3019.mp4, 3.035714286, 3.107142857, 3.214285714
+3021.mp4, 2.714285714, 3.314285714, 3.228571429
+3023.mp4, 3.028571429, 3.457142857, 3.228571429
+3029.mp4, 3.115384615, 3.115384615, 3.230769231
+3035.mp4, 3.192307692, 3.192307692, 3.230769231
+3036.mp4, 2.923076923, 3.230769231, 3.230769231
+3047.mp4, 3.230769231, 3.346153846, 3.230769231
+3049.mp4, 3.076923077, 3.346153846, 3.230769231
+3053.mp4, 3.153846154, 3.384615385, 3.230769231
+3065.mp4, 3.192307692, 3.538461538, 3.230769231
+3069.mp4, 3.076923077, 3.269230769, 3.230769231
+3074.mp4, 2.961538462, 3.576923077, 3.230769231
+3096.mp4, 3.0, 3.615384615, 3.230769231
+3101.mp4, 3.264705882, 3.470588235, 3.235294118
+3102.mp4, 3.294117647, 3.235294118, 3.235294118
+3103.mp4, 3.235294118, 3.264705882, 3.235294118
+3118.mp4, 3.206896552, 3.172413793, 3.24137931
+3123.mp4, 3.137931034, 3.344827586, 3.24137931
+3134.mp4, 3.107142857, 3.607142857, 3.25
+3138.mp4, 3.107142857, 3.428571429, 3.25
+3154.mp4, 3.117647059, 3.264705882, 3.264705882
+3157.mp4, 3.0, 3.147058824, 3.264705882
+3165.mp4, 3.230769231, 3.346153846, 3.269230769
+3170.mp4, 2.923076923, 3.423076923, 3.269230769
+3175.mp4, 2.961538462, 3.692307692, 3.269230769
+3182.mp4, 3.269230769, 3.269230769, 3.269230769
+3183.mp4, 3.192307692, 3.269230769, 3.269230769
+3192.mp4, 2.961538462, 3.423076923, 3.269230769
+3195.mp4, 3.192307692, 3.461538462, 3.269230769
+3196.mp4, 3.230769231, 3.5, 3.269230769
+3197.mp4, 3.038461538, 3.5, 3.269230769
+3200.mp4, 3.038461538, 3.653846154, 3.269230769
+3203.mp4, 3.038461538, 3.653846154, 3.269230769
+3204.mp4, 3.307692308, 3.115384615, 3.269230769
+3212.mp4, 3.192307692, 3.230769231, 3.269230769
+3213.mp4, 3.115384615, 3.307692308, 3.269230769
+3216.mp4, 3.034482759, 3.551724138, 3.275862069
+3221.mp4, 3.137931034, 3.344827586, 3.275862069
+3223.mp4, 3.137931034, 3.344827586, 3.275862069
+3228.mp4, 3.034482759, 3.517241379, 3.275862069
+3252.mp4, 3.057142857, 3.514285714, 3.285714286
+3253.mp4, 3.0, 3.542857143, 3.285714286
+3254.mp4, 2.942857143, 3.571428571, 3.285714286
+3257.mp4, 2.571428571, 3.742857143, 3.285714286
+3260.mp4, 3.176470588, 3.352941176, 3.294117647
+3262.mp4, 3.205882353, 3.411764706, 3.294117647
+3263.mp4, 3.205882353, 3.294117647, 3.294117647
+3270.mp4, 3.076923077, 3.307692308, 3.307692308
+3281.mp4, 3.115384615, 3.461538462, 3.307692308
+3287.mp4, 3.230769231, 3.538461538, 3.307692308
+3290.mp4, 2.961538462, 3.538461538, 3.307692308
+3291.mp4, 3.038461538, 3.576923077, 3.307692308
+3295.mp4, 3.5, 3.076923077, 3.307692308
+3301.mp4, 3.115384615, 3.576923077, 3.307692308
+3305.mp4, 3.307692308, 3.576923077, 3.307692308
+3313.mp4, 3.192307692, 3.5, 3.307692308
+3315.mp4, 3.230769231, 3.576923077, 3.307692308
+3317.mp4, 3.153846154, 3.692307692, 3.307692308
+3320.mp4, 3.103448276, 3.275862069, 3.310344828
+3327.mp4, 3.206896552, 3.24137931, 3.310344828
+3328.mp4, 3.068965517, 3.586206897, 3.310344828
+3336.mp4, 3.257142857, 3.342857143, 3.314285714
+3339.mp4, 3.171428571, 3.657142857, 3.314285714
+3340.mp4, 3.357142857, 3.321428571, 3.321428571
+3341.mp4, 3.214285714, 3.535714286, 3.321428571
+3343.mp4, 3.178571429, 3.392857143, 3.321428571
+3347.mp4, 3.178571429, 3.464285714, 3.321428571
+3348.mp4, 3.25, 3.357142857, 3.321428571
+3350.mp4, 3.178571429, 3.5, 3.321428571
+3351.mp4, 3.178571429, 3.428571429, 3.321428571
+3359.mp4, 3.382352941, 3.205882353, 3.323529412
+3362.mp4, 3.323529412, 3.147058824, 3.323529412
+3365.mp4, 3.085714286, 3.514285714, 3.342857143
+3383.mp4, 3.24137931, 3.413793103, 3.344827586
+3389.mp4, 3.379310345, 3.517241379, 3.344827586
+3392.mp4, 3.461538462, 3.307692308, 3.346153846
+3393.mp4, 3.230769231, 3.384615385, 3.346153846
+3396.mp4, 3.0, 3.461538462, 3.346153846
+3399.mp4, 3.423076923, 3.5, 3.346153846
+3404.mp4, 2.961538462, 3.576923077, 3.346153846
+3406.mp4, 3.307692308, 3.307692308, 3.346153846
+3409.mp4, 3.384615385, 3.346153846, 3.346153846
+3422.mp4, 3.192307692, 3.653846154, 3.346153846
+3427.mp4, 3.576923077, 3.0, 3.346153846
+3441.mp4, 3.384615385, 3.307692308, 3.346153846
+3447.mp4, 2.923076923, 3.576923077, 3.346153846
+3448.mp4, 3.264705882, 3.352941176, 3.352941176
+3455.mp4, 3.382352941, 3.411764706, 3.352941176
+3468.mp4, 3.071428571, 3.571428571, 3.357142857
+3473.mp4, 3.214285714, 3.464285714, 3.357142857
+3476.mp4, 3.142857143, 3.607142857, 3.357142857
+3481.mp4, 3.514285714, 3.514285714, 3.371428571
+3490.mp4, 3.310344828, 3.517241379, 3.379310345
+3493.mp4, 3.24137931, 3.379310345, 3.379310345
+3497.mp4, 3.275862069, 3.413793103, 3.379310345
+3498.mp4, 3.068965517, 3.310344828, 3.379310345
+3514.mp4, 3.352941176, 3.382352941, 3.382352941
+3519.mp4, 3.230769231, 3.307692308, 3.384615385
+3522.mp4, 3.230769231, 3.384615385, 3.384615385
+3530.mp4, 3.153846154, 3.5, 3.384615385
+3537.mp4, 3.192307692, 3.615384615, 3.384615385
+3542.mp4, 3.269230769, 3.346153846, 3.384615385
+3555.mp4, 3.230769231, 3.461538462, 3.384615385
+3556.mp4, 3.076923077, 3.5, 3.384615385
+3558.mp4, 3.269230769, 3.538461538, 3.384615385
+3570.mp4, 3.457142857, 3.171428571, 3.4
+3579.mp4, 3.411764706, 3.411764706, 3.411764706
+3580.mp4, 3.352941176, 3.5, 3.411764706
+3581.mp4, 3.470588235, 3.323529412, 3.411764706
+3582.mp4, 3.352941176, 3.617647059, 3.411764706
+3583.mp4, 3.323529412, 3.529411765, 3.411764706
+3593.mp4, 3.310344828, 3.586206897, 3.413793103
+3594.mp4, 3.482758621, 3.517241379, 3.413793103
+3595.mp4, 3.206896552, 3.75862069, 3.413793103
+3608.mp4, 3.068965517, 3.551724138, 3.413793103
+3610.mp4, 3.310344828, 3.586206897, 3.413793103
+3614.mp4, 3.192307692, 3.5, 3.423076923
+3620.mp4, 3.269230769, 3.615384615, 3.423076923
+3625.mp4, 3.307692308, 3.307692308, 3.423076923
+3636.mp4, 3.230769231, 3.5, 3.423076923
+3638.mp4, 3.307692308, 3.5, 3.423076923
+3642.mp4, 3.269230769, 3.576923077, 3.423076923
+3650.mp4, 3.230769231, 3.730769231, 3.423076923
+3655.mp4, 3.307692308, 3.653846154, 3.423076923
+3656.mp4, 3.346153846, 3.346153846, 3.423076923
+3658.mp4, 3.307692308, 3.538461538, 3.423076923
+3664.mp4, 3.5, 3.285714286, 3.428571429
+3676.mp4, 3.171428571, 3.714285714, 3.428571429
+3680.mp4, 3.382352941, 3.647058824, 3.441176471
+3686.mp4, 3.235294118, 3.470588235, 3.441176471
+3695.mp4, 3.310344828, 3.586206897, 3.448275862
+3696.mp4, 3.275862069, 3.586206897, 3.448275862
+3701.mp4, 3.551724138, 3.448275862, 3.448275862
+3704.mp4, 3.344827586, 3.448275862, 3.448275862
+3708.mp4, 2.942857143, 3.4, 3.457142857
+3710.mp4, 3.423076923, 3.423076923, 3.461538462
+3712.mp4, 3.269230769, 3.461538462, 3.461538462
+3717.mp4, 3.384615385, 3.576923077, 3.461538462
+3718.mp4, 3.346153846, 3.576923077, 3.461538462
+3720.mp4, 3.423076923, 3.615384615, 3.461538462
+3721.mp4, 3.153846154, 3.615384615, 3.461538462
+3722.mp4, 3.307692308, 3.615384615, 3.461538462
+3728.mp4, 3.384615385, 3.5, 3.461538462
+3729.mp4, 3.5, 3.346153846, 3.461538462
+3732.mp4, 3.076923077, 3.538461538, 3.461538462
+3738.mp4, 3.076923077, 3.884615385, 3.461538462
+3740.mp4, 3.346153846, 3.384615385, 3.461538462
+3742.mp4, 3.384615385, 3.461538462, 3.461538462
+3759.mp4, 3.5, 3.5, 3.464285714
+3761.mp4, 3.285714286, 3.678571429, 3.464285714
+3762.mp4, 3.323529412, 3.529411765, 3.470588235
+3764.mp4, 3.441176471, 3.735294118, 3.470588235
+3765.mp4, 3.382352941, 3.441176471, 3.470588235
+3767.mp4, 3.5, 3.441176471, 3.470588235
+3771.mp4, 3.482758621, 3.551724138, 3.482758621
+3772.mp4, 3.413793103, 3.586206897, 3.482758621
+3781.mp4, 3.413793103, 3.413793103, 3.482758621
+3783.mp4, 3.413793103, 3.344827586, 3.482758621
+3785.mp4, 3.448275862, 3.275862069, 3.482758621
+3790.mp4, 3.230769231, 3.307692308, 3.5
+3792.mp4, 3.384615385, 3.384615385, 3.5
+3795.mp4, 3.615384615, 3.423076923, 3.5
+3797.mp4, 3.423076923, 3.423076923, 3.5
+3799.mp4, 3.230769231, 3.461538462, 3.5
+3814.mp4, 3.192307692, 3.615384615, 3.5
+3815.mp4, 3.384615385, 3.615384615, 3.5
+3818.mp4, 3.5, 3.615384615, 3.5
+3821.mp4, 3.107142857, 3.678571429, 3.5
+3826.mp4, 3.346153846, 3.807692308, 3.5
+3829.mp4, 3.730769231, 3.423076923, 3.5
+3836.mp4, 3.307692308, 3.307692308, 3.5
+3838.mp4, 3.321428571, 3.571428571, 3.5
+3839.mp4, 3.25, 3.821428571, 3.5
+3845.mp4, 3.529411765, 3.382352941, 3.5
+3860.mp4, 3.413793103, 3.482758621, 3.517241379
+3866.mp4, 3.482758621, 3.586206897, 3.517241379
+3871.mp4, 3.588235294, 3.558823529, 3.529411765
+3878.mp4, 3.464285714, 3.392857143, 3.535714286
+3880.mp4, 3.307692308, 3.538461538, 3.538461538
+3883.mp4, 3.538461538, 3.538461538, 3.538461538
+3884.mp4, 3.384615385, 3.576923077, 3.538461538
+3885.mp4, 3.384615385, 3.576923077, 3.538461538
+3886.mp4, 3.307692308, 3.615384615, 3.538461538
+3888.mp4, 3.384615385, 3.653846154, 3.538461538
+3895.mp4, 3.576923077, 3.461538462, 3.538461538
+3904.mp4, 3.461538462, 3.653846154, 3.538461538
+3908.mp4, 3.346153846, 3.730769231, 3.538461538
+3909.mp4, 3.461538462, 3.769230769, 3.538461538
+3918.mp4, 3.342857143, 3.542857143, 3.542857143
+3925.mp4, 3.482758621, 3.620689655, 3.551724138
+3928.mp4, 3.344827586, 3.827586207, 3.551724138
+3930.mp4, 3.413793103, 3.551724138, 3.551724138
+3948.mp4, 3.285714286, 3.857142857, 3.571428571
+3951.mp4, 3.428571429, 3.607142857, 3.571428571
+3952.mp4, 3.428571429, 3.678571429, 3.571428571
+3954.mp4, 3.5, 3.538461538, 3.576923077
+3957.mp4, 3.461538462, 3.615384615, 3.576923077
+3959.mp4, 3.5, 3.615384615, 3.576923077
+3963.mp4, 3.730769231, 3.692307692, 3.576923077
+3967.mp4, 3.538461538, 3.884615385, 3.576923077
+3971.mp4, 3.5, 3.730769231, 3.576923077
+3977.mp4, 3.192307692, 3.423076923, 3.576923077
+3991.mp4, 3.551724138, 3.551724138, 3.586206897
+3992.mp4, 3.379310345, 3.827586207, 3.586206897
+3995.mp4, 3.586206897, 3.517241379, 3.586206897
+4009.mp4, 3.285714286, 3.571428571, 3.6
+4018.mp4, 3.571428571, 3.642857143, 3.607142857
+4020.mp4, 3.576923077, 3.576923077, 3.615384615
+4022.mp4, 3.346153846, 3.615384615, 3.615384615
+4026.mp4, 3.615384615, 3.730769231, 3.615384615
+4031.mp4, 3.384615385, 3.5, 3.615384615
+4035.mp4, 3.346153846, 3.653846154, 3.615384615
+4037.mp4, 3.461538462, 3.653846154, 3.615384615
+4054.mp4, 3.275862069, 3.655172414, 3.620689655
+4057.mp4, 3.517241379, 3.620689655, 3.620689655
+4063.mp4, 3.551724138, 3.620689655, 3.620689655
+4064.mp4, 3.413793103, 3.448275862, 3.620689655
+4069.mp4, 3.571428571, 3.742857143, 3.628571429
+4072.mp4, 3.678571429, 3.642857143, 3.642857143
+4073.mp4, 3.714285714, 3.75, 3.642857143
+4078.mp4, 3.588235294, 3.588235294, 3.647058824
+4084.mp4, 3.576923077, 3.653846154, 3.653846154
+4087.mp4, 3.576923077, 3.692307692, 3.653846154
+4091.mp4, 3.615384615, 3.769230769, 3.653846154
+4098.mp4, 3.5, 3.615384615, 3.653846154
+4103.mp4, 3.576923077, 3.615384615, 3.653846154
+4105.mp4, 3.923076923, 3.692307692, 3.653846154
+4115.mp4, 3.551724138, 3.896551724, 3.655172414
+4120.mp4, 3.448275862, 3.793103448, 3.655172414
+4123.mp4, 3.344827586, 3.827586207, 3.655172414
+4125.mp4, 3.724137931, 3.586206897, 3.655172414
+4131.mp4, 3.535714286, 3.892857143, 3.678571429
+4132.mp4, 3.714285714, 3.964285714, 3.678571429
+4133.mp4, 3.464285714, 3.75, 3.678571429
+4142.mp4, 3.551724138, 3.655172414, 3.689655172
+4148.mp4, 3.517241379, 3.724137931, 3.689655172
+4149.mp4, 3.586206897, 3.862068966, 3.689655172
+4151.mp4, 3.653846154, 3.615384615, 3.692307692
+4152.mp4, 3.307692308, 3.692307692, 3.692307692
+4154.mp4, 3.538461538, 3.807692308, 3.692307692
+4156.mp4, 3.5, 3.846153846, 3.692307692
+4183.mp4, 3.371428571, 3.771428571, 3.714285714
+4187.mp4, 3.457142857, 3.914285714, 3.714285714
+4218.mp4, 3.692307692, 3.884615385, 3.730769231
+4222.mp4, 3.4, 3.857142857, 3.742857143
+4224.mp4, 3.678571429, 3.607142857, 3.75
+4228.mp4, 3.642857143, 3.75, 3.75
+4230.mp4, 3.655172414, 3.75862069, 3.75862069
+4231.mp4, 3.75862069, 3.793103448, 3.75862069
+4234.mp4, 3.586206897, 3.965517241, 3.75862069
+4237.mp4, 3.676470588, 3.794117647, 3.764705882
+4243.mp4, 3.576923077, 3.846153846, 3.769230769
+4244.mp4, 3.538461538, 3.846153846, 3.769230769
+4245.mp4, 3.576923077, 3.884615385, 3.769230769
+4254.mp4, 3.692307692, 3.846153846, 3.769230769
+4265.mp4, 3.807692308, 3.846153846, 3.769230769
+4269.mp4, 3.457142857, 4.0, 3.771428571
+4274.mp4, 3.642857143, 3.964285714, 3.785714286
+4281.mp4, 3.793103448, 4.0, 3.793103448
+4284.mp4, 3.793103448, 3.862068966, 3.793103448
+4288.mp4, 3.911764706, 3.764705882, 3.794117647
+4290.mp4, 3.485714286, 3.942857143, 3.8
+4294.mp4, 3.730769231, 3.769230769, 3.807692308
+4303.mp4, 3.730769231, 3.769230769, 3.807692308
+4311.mp4, 3.620689655, 3.827586207, 3.827586207
+4312.mp4, 3.896551724, 3.620689655, 3.827586207
+4314.mp4, 3.862068966, 3.724137931, 3.827586207
+4315.mp4, 3.655172414, 3.896551724, 3.827586207
+4316.mp4, 3.714285714, 3.571428571, 3.828571429
+4318.mp4, 3.885714286, 4.085714286, 3.828571429
+4333.mp4, 3.642857143, 3.892857143, 3.857142857
+4338.mp4, 3.642857143, 3.928571429, 3.857142857
+4343.mp4, 3.793103448, 3.724137931, 3.862068966
+4353.mp4, 3.730769231, 3.884615385, 3.884615385
+4356.mp4, 3.769230769, 4.0, 3.884615385
+4361.mp4, 3.896551724, 3.896551724, 3.896551724
+4362.mp4, 3.823529412, 3.852941176, 3.911764706
+4364.mp4, 3.882352941, 3.882352941, 3.911764706
+4370.mp4, 4.0, 3.846153846, 3.923076923
+4377.mp4, 3.961538462, 4.0, 3.923076923
+4380.mp4, 3.961538462, 3.961538462, 3.923076923
+4382.mp4, 3.75, 3.892857143, 3.928571429
+4386.mp4, 3.896551724, 3.931034483, 3.931034483
+4388.mp4, 3.793103448, 3.862068966, 3.931034483
+4389.mp4, 3.896551724, 3.827586207, 3.931034483
+4391.mp4, 3.735294118, 3.970588235, 3.941176471
+4393.mp4, 3.923076923, 4.0, 3.961538462
+4412.mp4, 3.961538462, 3.961538462, 4.0
+4413.mp4, 4.0, 3.961538462, 4.0
+4414.mp4, 3.714285714, 3.964285714, 4.0
+4418.mp4, 3.923076923, 4.038461538, 4.0
+4420.mp4, 3.892857143, 4.071428571, 4.0
+4421.mp4, 3.923076923, 4.115384615, 4.0
+4423.mp4, 3.961538462, 4.0, 4.0
+4426.mp4, 3.724137931, 3.896551724, 4.0
+4427.mp4, 4.068965517, 3.965517241, 4.0
+4433.mp4, 4.0, 4.034482759, 4.034482759
+4437.mp4, 4.035714286, 4.071428571, 4.035714286
+4438.mp4, 4.038461538, 4.115384615, 4.038461538
+4439.mp4, 3.961538462, 4.0, 4.038461538
+4448.mp4, 4.107142857, 4.142857143, 4.071428571
+4454.mp4, 4.115384615, 4.115384615, 4.076923077
+4455.mp4, 4.076923077, 4.307692308, 4.076923077
+4461.mp4, 4.034482759, 3.896551724, 4.103448276
+4462.mp4, 3.862068966, 4.137931034, 4.103448276
+4463.mp4, 3.785714286, 4.142857143, 4.107142857
+4469.mp4, 4.076923077, 4.038461538, 4.115384615
+4470.mp4, 3.884615385, 4.076923077, 4.115384615
+4474.mp4, 4.076923077, 4.230769231, 4.115384615
+4477.mp4, 4.103448276, 4.206896552, 4.137931034
+4482.mp4, 4.0, 3.961538462, 4.153846154
+4484.mp4, 4.307692308, 4.153846154, 4.153846154
+4497.mp4, 4.172413793, 4.172413793, 4.206896552
+4504.mp4, 4.071428571, 4.428571429, 4.25
+4517.mp4, 4.068965517, 4.379310345, 4.310344828
+4521.mp4, 4.230769231, 4.307692308, 4.346153846
+4523.mp4, 4.269230769, 4.461538462, 4.346153846
+4524.mp4, 4.142857143, 4.5, 4.357142857
+4526.mp4, 4.24137931, 4.517241379, 4.379310345
+4527.mp4, 4.307692308, 4.384615385, 4.384615385
+4535.mp4, 4.314285714, 4.4, 4.428571429
+4536.mp4, 4.171428571, 4.4, 4.457142857
+4538.mp4, 4.428571429, 4.428571429, 4.464285714
+4541.mp4, 4.4, 4.514285714, 4.685714286

examplar_data_labels/KoNViD/labels.txt

@@ -0,0 +1,1200 @@
+KoNViD_1k_videos/4542323058.mp4, 8.008, 29.97002997002997, 3.22
+KoNViD_1k_videos/9753414792.mp4, 8.008008, 29.97, 3.84
+KoNViD_1k_videos/6935410837.mp4, 8.0, 25.0, 3.24
+KoNViD_1k_videos/8171831850.mp4, 8.008, 29.97002997002997, 4.444029851
+KoNViD_1k_videos/11465976586.mp4, 8.008008, 29.97, 3.94
+KoNViD_1k_videos/4323977167.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/7005478889.mp4, 8.008, 23.976023976023978, 2.86
+KoNViD_1k_videos/6346026937.mp4, 8.008, 23.976023976023978, 2.14
+KoNViD_1k_videos/10672253555.mp4, 8.008008, 29.97, 1.614785992
+KoNViD_1k_videos/6812062788.mp4, 8.008, 29.97002997002997, 2.52
+KoNViD_1k_videos/3930579113.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/6981897920.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/8957212602.mp4, 8.008008, 29.97, 3.56
+KoNViD_1k_videos/4771991539.mp4, 8.0, 24.0, 3.0
+KoNViD_1k_videos/8420230247.mp4, 8.008, 29.97002997002997, 3.96
+KoNViD_1k_videos/4161940753.mp4, 8.0, 24.0, 3.66
+KoNViD_1k_videos/8408744905.mp4, 8.008, 29.97002997002997, 3.5
+KoNViD_1k_videos/13291575674.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/6538025379.mp4, 8.008, 29.97002997002997, 3.8
+KoNViD_1k_videos/7344072960.mp4, 8.0, 24.0, 2.769874477
+KoNViD_1k_videos/9969637164.mp4, 8.008008, 29.97, 3.42
+KoNViD_1k_videos/8688311915.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/6929069239.mp4, 8.008, 29.97002997002997, 2.58
+KoNViD_1k_videos/9248029519.mp4, 8.008008, 29.97, 2.7
+KoNViD_1k_videos/3617102785.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/9189459433.mp4, 8.0, 25.0, 2.22
+KoNViD_1k_videos/3359662128.mp4, 8.008, 23.976023976023978, 3.08
+KoNViD_1k_videos/6408325533.mp4, 8.008, 29.97002997002997, 1.26
+KoNViD_1k_videos/5310216885.mp4, 8.008, 29.97002997002997, 2.24
+KoNViD_1k_videos/9455400456.mp4, 8.0, 24.0, 2.22
+KoNViD_1k_videos/7820483368.mp4, 8.008, 29.97002997002997, 2.76
+KoNViD_1k_videos/4713503488.mp4, 8.008, 23.976023976023978, 3.88
+KoNViD_1k_videos/6024650797.mp4, 8.008, 29.97002997002997, 2.54
+KoNViD_1k_videos/8548046708.mp4, 8.008, 23.976023976023978, 3.36
+KoNViD_1k_videos/8078127133.mp4, 8.008, 29.97002997002997, 2.618867925
+KoNViD_1k_videos/5490599661.mp4, 8.008, 29.97002997002997, 2.88
+KoNViD_1k_videos/6749780445.mp4, 8.008, 29.97002997002997, 2.02
+KoNViD_1k_videos/5433352304.mp4, 8.008, 29.97002997002997, 2.22
+KoNViD_1k_videos/12262452854.mp4, 8.0, 24.0, 4.24
+KoNViD_1k_videos/4299787720.mp4, 8.008, 29.97002997002997, 2.849802372
+KoNViD_1k_videos/4453069241.mp4, 8.008, 29.97002997002997, 2.92
+KoNViD_1k_videos/4915461487.mp4, 8.0, 24.0, 2.24
+KoNViD_1k_videos/8565847545.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/7012065923.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/4970919913.mp4, 8.0, 24.0, 2.32
+KoNViD_1k_videos/5424989067.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/3318634083.mp4, 8.008, 29.97002997002997, 2.3
+KoNViD_1k_videos/8317937019.mp4, 8.008, 29.97002997002997, 2.48
+KoNViD_1k_videos/8667534666.mp4, 8.008, 29.97002997002997, 4.02
+KoNViD_1k_videos/6060761361.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/4161747134.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/8937776831.mp4, 8.008008, 29.97, 3.9
+KoNViD_1k_videos/11237101334.mp4, 8.0, 24.0, 2.26
+KoNViD_1k_videos/8182593817.mp4, 8.008, 29.97002997002997, 2.8
+KoNViD_1k_videos/11403577283.mp4, 8.008, 23.976023976023978, 3.06
+KoNViD_1k_videos/4482595368.mp4, 8.008, 29.97002997002997, 3.86
+KoNViD_1k_videos/5564650233.mp4, 8.008, 29.97002997002997, 3.484251969
+KoNViD_1k_videos/4626131441.mp4, 8.008, 23.976023976023978, 2.68
+KoNViD_1k_videos/6100104774.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/8088901520.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/4557456762.mp4, 8.008, 29.97002997002997, 3.24
+KoNViD_1k_videos/5622498540.mp4, 8.008, 29.97002997002997, 3.0
+KoNViD_1k_videos/9086621277.mp4, 8.0, 24.0, 3.28
+KoNViD_1k_videos/8097209231.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/5263204150.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/3407992311.mp4, 8.008, 29.97002997002997, 4.08
+KoNViD_1k_videos/4472493392.mp4, 8.008, 23.976023976023978, 2.38
+KoNViD_1k_videos/6350328925.mp4, 8.008, 23.976023976023978, 3.4
+KoNViD_1k_videos/5512609909.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/6047579133.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/8592011515.mp4, 8.0, 24.0, 2.0
+KoNViD_1k_videos/7834396986.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/8554748075.mp4, 8.008, 23.976023976023978, 2.089068826
+KoNViD_1k_videos/6073473868.mp4, 8.0, 24.0, 2.62
+KoNViD_1k_videos/3887621073.mp4, 8.008, 29.97002997002997, 1.58
+KoNViD_1k_videos/6874423527.mp4, 8.0, 24.0, 2.82
+KoNViD_1k_videos/7123482247.mp4, 8.008, 23.976023976023978, 3.258064516
+KoNViD_1k_videos/6008700446.mp4, 8.008, 29.97002997002997, 2.0
+KoNViD_1k_videos/12816380173.mp4, 8.008, 23.976023976023978, 3.28
+KoNViD_1k_videos/3504083604.mp4, 8.008, 29.97002997002997, 2.38
+KoNViD_1k_videos/5962679589.mp4, 8.008, 29.97002997002997, 1.86
+KoNViD_1k_videos/8236900278.mp4, 8.0, 24.0, 3.36
+KoNViD_1k_videos/8909265380.mp4, 8.008008, 29.97, 2.78
+KoNViD_1k_videos/3015973424.mp4, 8.008, 23.976023976023978, 3.1
+KoNViD_1k_videos/11646665363.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/8151400613.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/6830500351.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/6680559525.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/5858864133.mp4, 8.008, 29.97002997002997, 2.48
+KoNViD_1k_videos/7175971942.mp4, 8.008, 29.97002997002997, 2.32
+KoNViD_1k_videos/3664548641.mp4, 8.0, 24.0, 3.121568627
+KoNViD_1k_videos/10208239954.mp4, 8.008008, 29.97, 3.12
+KoNViD_1k_videos/8773297527.mp4, 8.008008, 29.97, 3.72
+KoNViD_1k_videos/8915047035.mp4, 8.008008, 29.97, 3.68
+KoNViD_1k_videos/10602272313.mp4, 8.008008, 29.97, 3.1
+KoNViD_1k_videos/5621374631.mp4, 8.008, 29.97002997002997, 2.52
+KoNViD_1k_videos/9732912473.mp4, 8.008008, 29.97, 3.42
+KoNViD_1k_videos/4991520511.mp4, 8.008, 29.97002997002997, 1.98
+KoNViD_1k_videos/6811275923.mp4, 8.008, 23.976023976023978, 3.36
+KoNViD_1k_videos/5169144673.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/5790468384.mp4, 8.0, 24.0, 2.36
+KoNViD_1k_videos/6571028091.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/5920289173.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/5032172516.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/3715429501.mp4, 8.008, 29.97002997002997, 2.9
+KoNViD_1k_videos/4970624827.mp4, 8.0, 25.0, 1.66
+KoNViD_1k_videos/7521145978.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/5825661459.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/8479364057.mp4, 8.008, 23.976023976023978, 2.74
+KoNViD_1k_videos/8148981356.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/12567402335.mp4, 8.008008, 29.97, 1.8
+KoNViD_1k_videos/8589300397.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/10500520233.mp4, 8.008008, 29.97, 2.86
+KoNViD_1k_videos/3934294391.mp4, 8.008, 29.97002997002997, 2.26
+KoNViD_1k_videos/8387302863.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/7402519316.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/7651248048.mp4, 8.0, 24.0, 2.08
+KoNViD_1k_videos/4005506137.mp4, 8.008, 23.976023976023978, 2.24
+KoNViD_1k_videos/3459410764.mp4, 8.008, 29.97002997002997, 3.692946058
+KoNViD_1k_videos/8039356416.mp4, 8.008, 29.97002997002997, 2.46
+KoNViD_1k_videos/5698463243.mp4, 8.0, 25.0, 2.96
+KoNViD_1k_videos/6865667129.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/4956991827.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/4879585643.mp4, 8.008, 29.97002997002997, 1.84
+KoNViD_1k_videos/8267321210.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/8240363350.mp4, 8.008, 29.97002997002997, 3.22
+KoNViD_1k_videos/5590704113.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/5045897151.mp4, 8.008, 29.97002997002997, 2.24
+KoNViD_1k_videos/6178882861.mp4, 8.008, 29.97002997002997, 2.76
+KoNViD_1k_videos/7068941889.mp4, 8.008, 29.97002997002997, 2.8
+KoNViD_1k_videos/3459287580.mp4, 8.008, 23.976023976023978, 3.5
+KoNViD_1k_videos/4593126957.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/8562619627.mp4, 8.0, 25.0, 4.26
+KoNViD_1k_videos/13036222674.mp4, 8.008008, 29.97, 3.56
+KoNViD_1k_videos/11540616946.mp4, 8.008008, 29.97, 4.08
+KoNViD_1k_videos/4979493912.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/12859019173.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/5222599467.mp4, 8.008, 29.97002997002997, 2.6
+KoNViD_1k_videos/6809728071.mp4, 8.008, 29.97002997002997, 3.52
+KoNViD_1k_videos/5075205526.mp4, 8.0, 24.0, 3.18
+KoNViD_1k_videos/6783810490.mp4, 8.008, 29.97002997002997, 1.88
+KoNViD_1k_videos/8230249635.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/8092884510.mp4, 8.0, 24.0, 2.42
+KoNViD_1k_videos/8625682550.mp4, 8.0, 24.0, 1.84
+KoNViD_1k_videos/8717647202.mp4, 8.0, 25.0, 3.64
+KoNViD_1k_videos/3337642103.mp4, 8.008, 29.97002997002997, 3.308641975
+KoNViD_1k_videos/8931854960.mp4, 8.008008, 29.97, 3.7
+KoNViD_1k_videos/9459232387.mp4, 8.008008, 29.97, 3.04
+KoNViD_1k_videos/4866009712.mp4, 8.008, 29.97002997002997, 2.42
+KoNViD_1k_videos/8679156964.mp4, 8.0, 25.0, 3.8
+KoNViD_1k_videos/8372557828.mp4, 8.008, 29.97002997002997, 2.08
+KoNViD_1k_videos/8591330227.mp4, 8.0, 25.0, 2.36
+KoNViD_1k_videos/6173852301.mp4, 8.008, 29.97002997002997, 2.2
+KoNViD_1k_videos/5620430600.mp4, 8.008, 29.97002997002997, 2.5
+KoNViD_1k_videos/12412571563.mp4, 8.008008, 29.97, 3.5
+KoNViD_1k_videos/8689455432.mp4, 8.008, 29.97002997002997, 2.68
+KoNViD_1k_videos/9427595515.mp4, 8.008008, 29.97, 2.06
+KoNViD_1k_videos/5379891328.mp4, 8.008, 29.97002997002997, 3.5
+KoNViD_1k_videos/5596033251.mp4, 8.008, 23.976023976023978, 3.66
+KoNViD_1k_videos/9387089233.mp4, 8.008008, 29.97, 2.68
+KoNViD_1k_videos/7547384258.mp4, 8.008, 29.97002997002997, 2.44
+KoNViD_1k_videos/7708735292.mp4, 8.0, 24.0, 2.46
+KoNViD_1k_videos/3533765114.mp4, 8.0, 24.0, 2.28
+KoNViD_1k_videos/5280283886.mp4, 8.008, 29.97002997002997, 3.86
+KoNViD_1k_videos/10655599784.mp4, 8.008008, 29.97, 3.26
+KoNViD_1k_videos/3321308714.mp4, 8.008, 23.976023976023978, 3.04
+KoNViD_1k_videos/6953979425.mp4, 8.008, 29.97002997002997, 3.02
+KoNViD_1k_videos/4095112467.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/9521557660.mp4, 8.0, 24.0, 3.6
+KoNViD_1k_videos/5111411754.mp4, 8.008, 29.97002997002997, 2.68
+KoNViD_1k_videos/5072411175.mp4, 8.0, 24.0, 2.5
+KoNViD_1k_videos/3589925544.mp4, 8.0, 24.0, 3.16
+KoNViD_1k_videos/7123683275.mp4, 8.0, 24.0, 2.06
+KoNViD_1k_videos/7074513729.mp4, 8.008, 29.97002997002997, 1.56
+KoNViD_1k_videos/12465909365.mp4, 8.008008, 29.97, 3.88
+KoNViD_1k_videos/3240926995.mp4, 8.008, 29.97002997002997, 3.1484375
+KoNViD_1k_videos/12952440214.mp4, 8.008, 29.97002997002997, 2.5
+KoNViD_1k_videos/7236880666.mp4, 8.008, 29.97002997002997, 3.5
+KoNViD_1k_videos/9144368303.mp4, 8.008008, 23.976, 2.9
+KoNViD_1k_videos/10864372176.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/7171297739.mp4, 8.008, 23.976023976023978, 2.06
+KoNViD_1k_videos/3561516939.mp4, 8.008, 29.97002997002997, 2.44
+KoNViD_1k_videos/4820762619.mp4, 8.008, 29.97002997002997, 2.16
+KoNViD_1k_videos/6868110990.mp4, 8.008, 29.97002997002997, 3.18
+KoNViD_1k_videos/7173586391.mp4, 8.008, 23.976023976023978, 2.74
+KoNViD_1k_videos/4391946048.mp4, 7.5742329999999995, 29.97002997002997, 3.3
+KoNViD_1k_videos/7274180028.mp4, 8.008, 23.976023976023978, 3.76
+KoNViD_1k_videos/6062892973.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/5149271412.mp4, 8.008, 29.97002997002997, 2.82
+KoNViD_1k_videos/7117529219.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/6937597827.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/12355531285.mp4, 8.008008, 29.97, 4.22
+KoNViD_1k_videos/5266562897.mp4, 8.008, 29.97002997002997, 2.56
+KoNViD_1k_videos/5921571324.mp4, 8.008, 29.97002997002997, 3.24
+KoNViD_1k_videos/8478531549.mp4, 8.008, 29.97002997002997, 2.3
+KoNViD_1k_videos/6422437813.mp4, 8.008, 29.97002997002997, 2.46
+KoNViD_1k_videos/5215773911.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/13276560083.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/6806699713.mp4, 8.008, 29.97002997002997, 3.76
+KoNViD_1k_videos/8257892715.mp4, 8.008, 29.97002997002997, 1.98
+KoNViD_1k_videos/7879889838.mp4, 8.008, 29.97002997002997, 2.76
+KoNViD_1k_videos/5218188823.mp4, 8.0, 24.0, 3.0
+KoNViD_1k_videos/5044148124.mp4, 8.008, 29.97002997002997, 2.18
+KoNViD_1k_videos/5917982116.mp4, 8.008, 29.97002997002997, 3.84
+KoNViD_1k_videos/5143368697.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/7390590300.mp4, 8.008, 23.976023976023978, 2.28
+KoNViD_1k_videos/6866970306.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/5026138416.mp4, 8.008, 29.97002997002997, 3.22
+KoNViD_1k_videos/5268797415.mp4, 8.0, 24.0, 2.44
+KoNViD_1k_videos/12965735905.mp4, 8.008008, 29.97, 3.18
+KoNViD_1k_videos/4800261002.mp4, 8.008, 29.97002997002997, 3.98
+KoNViD_1k_videos/9513714022.mp4, 8.008008, 29.97, 3.8
+KoNViD_1k_videos/7495255120.mp4, 8.008, 23.976023976023978, 2.42
+KoNViD_1k_videos/5223525504.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/10127250986.mp4, 8.008008, 29.97, 2.3
+KoNViD_1k_videos/6971905222.mp4, 8.008, 29.97002997002997, 2.6
+KoNViD_1k_videos/6250048696.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/12044525214.mp4, 8.008008, 29.97, 3.0
+KoNViD_1k_videos/6181812275.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/10716218856.mp4, 8.008008, 23.976, 2.06
+KoNViD_1k_videos/6724048499.mp4, 8.008, 29.97002997002997, 2.56
+KoNViD_1k_videos/12379836133.mp4, 8.0, 25.0, 4.24
+KoNViD_1k_videos/4877128768.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/10609570083.mp4, 8.008008, 29.97, 3.1
+KoNViD_1k_videos/5319047612.mp4, 8.008, 29.97002997002997, 1.346863469
+KoNViD_1k_videos/9000209228.mp4, 8.0, 25.0, 2.1
+KoNViD_1k_videos/12923454185.mp4, 8.008, 29.97002997002997, 2.88
+KoNViD_1k_videos/6868246166.mp4, 8.0, 24.0, 2.68
+KoNViD_1k_videos/4397381332.mp4, 8.008, 29.97002997002997, 3.56
+KoNViD_1k_videos/7381539666.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/4971091773.mp4, 8.0, 24.0, 2.9
+KoNViD_1k_videos/4655662252.mp4, 8.008, 29.97002997002997, 3.94
+KoNViD_1k_videos/5947649803.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/4375496636.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/4980230630.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/8140107150.mp4, 8.0, 24.0, 2.4
+KoNViD_1k_videos/8017059368.mp4, 8.008, 29.97002997002997, 1.907172996
+KoNViD_1k_videos/5118641893.mp4, 8.0, 24.0, 2.76
+KoNViD_1k_videos/7313643370.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/5542380796.mp4, 7.708333, 24.0, 1.72
+KoNViD_1k_videos/4565232312.mp4, 8.008, 29.97002997002997, 2.64
+KoNViD_1k_videos/4573266245.mp4, 8.008, 29.97002997002997, 2.58
+KoNViD_1k_videos/4959664675.mp4, 8.0, 24.0, 2.06
+KoNViD_1k_videos/6370967061.mp4, 8.0, 24.0, 2.2
+KoNViD_1k_videos/7954124226.mp4, 8.008, 29.97002997002997, 3.8
+KoNViD_1k_videos/6096480589.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/8069024908.mp4, 8.008, 29.97002997002997, 2.12
+KoNViD_1k_videos/6971190423.mp4, 8.008, 29.97002997002997, 3.82
+KoNViD_1k_videos/4477789338.mp4, 8.0, 24.0, 2.72
+KoNViD_1k_videos/9258419063.mp4, 8.0, 25.0, 3.74
+KoNViD_1k_videos/8657921837.mp4, 8.0, 24.0, 4.3
+KoNViD_1k_videos/8477990214.mp4, 8.0, 24.0, 2.1
+KoNViD_1k_videos/7978973307.mp4, 8.008, 29.97002997002997, 1.78
+KoNViD_1k_videos/13079468475.mp4, 8.008008, 29.97, 3.68
+KoNViD_1k_videos/5694745216.mp4, 8.0, 24.0, 2.66
+KoNViD_1k_videos/8495275628.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/9765162404.mp4, 8.008008, 29.97, 2.08
+KoNViD_1k_videos/4642629203.mp4, 8.008, 23.976023976023978, 3.52
+KoNViD_1k_videos/10270536464.mp4, 8.008008, 29.97, 3.42
+KoNViD_1k_videos/4914377539.mp4, 8.008, 29.97002997002997, 2.2
+KoNViD_1k_videos/5493602442.mp4, 8.0, 24.0, 2.5
+KoNViD_1k_videos/5737374844.mp4, 8.0, 24.0, 3.0
+KoNViD_1k_videos/8072548302.mp4, 8.008, 29.97002997002997, 3.02
+KoNViD_1k_videos/3357588876.mp4, 8.0, 25.0, 2.96
+KoNViD_1k_videos/6932519097.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/4615948555.mp4, 8.008, 23.976023976023978, 3.02
+KoNViD_1k_videos/7146539585.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/5322533709.mp4, 8.008, 29.97002997002997, 2.76
+KoNViD_1k_videos/5823644068.mp4, 8.0, 24.0, 2.9
+KoNViD_1k_videos/12758594775.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/7246583832.mp4, 8.0, 24.0, 2.86
+KoNViD_1k_videos/5995204968.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/6313200205.mp4, 8.008, 29.97002997002997, 3.78
+KoNViD_1k_videos/4966446054.mp4, 8.0, 24.0, 2.26
+KoNViD_1k_videos/9740517228.mp4, 8.008008, 23.976, 3.8
+KoNViD_1k_videos/3863745355.mp4, 8.008, 23.976023976023978, 4.02
+KoNViD_1k_videos/5902098420.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/8483400755.mp4, 8.008, 29.97002997002997, 1.46
+KoNViD_1k_videos/5579073175.mp4, 8.0, 25.0, 3.32
+KoNViD_1k_videos/5756899543.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/5831253770.mp4, 8.008, 29.97002997002997, 3.84
+KoNViD_1k_videos/6492110563.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/9709640314.mp4, 8.008008, 29.97, 4.02
+KoNViD_1k_videos/5263261013.mp4, 8.008, 29.97002997002997, 2.74
+KoNViD_1k_videos/6812860936.mp4, 8.008, 29.97002997002997, 2.88
+KoNViD_1k_videos/8223463984.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/4779773181.mp4, 8.0, 25.0, 4.14
+KoNViD_1k_videos/12816958564.mp4, 8.0, 25.0, 2.84
+KoNViD_1k_videos/5583994811.mp4, 8.008, 29.97002997002997, 3.529166667
+KoNViD_1k_videos/4866613201.mp4, 8.008, 29.97002997002997, 2.2
+KoNViD_1k_videos/4729578673.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/4260295122.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/4620516469.mp4, 8.008, 23.976023976023978, 3.52
+KoNViD_1k_videos/7700672030.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/8281263977.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/9525693928.mp4, 8.008008, 29.97, 3.86
+KoNViD_1k_videos/5858312093.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/4136131472.mp4, 8.0, 24.0, 3.2
+KoNViD_1k_videos/10521676086.mp4, 8.008008, 29.97, 2.88
+KoNViD_1k_videos/5031142078.mp4, 8.0, 24.0, 2.2
+KoNViD_1k_videos/12626612584.mp4, 8.008008, 29.97, 3.1
+KoNViD_1k_videos/8069868772.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/5816132223.mp4, 8.008, 29.97002997002997, 3.58
+KoNViD_1k_videos/5052341075.mp4, 8.008, 23.976023976023978, 3.48
+KoNViD_1k_videos/4029116553.mp4, 8.008, 23.976023976023978, 3.82
+KoNViD_1k_videos/6894958520.mp4, 8.008, 29.97002997002997, 3.24
+KoNViD_1k_videos/3471567822.mp4, 8.008, 29.97002997002997, 1.52
+KoNViD_1k_videos/11920895013.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/8491617830.mp4, 8.008, 23.976023976023978, 3.5
+KoNViD_1k_videos/6948249567.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/7832506570.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/10541657773.mp4, 8.0, 24.0, 3.74
+KoNViD_1k_videos/3860380907.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/8286421427.mp4, 8.0, 25.0, 4.14
+KoNViD_1k_videos/10694139655.mp4, 8.008008, 23.976, 1.82
+KoNViD_1k_videos/8480588816.mp4, 8.008, 29.97002997002997, 3.74
+KoNViD_1k_videos/9089204842.mp4, 8.008008, 23.976, 3.02
+KoNViD_1k_videos/5412442677.mp4, 8.008, 29.97002997002997, 2.3
+KoNViD_1k_videos/4465669054.mp4, 8.0, 24.0, 2.92
+KoNViD_1k_videos/6059786344.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/13286356624.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/4842303128.mp4, 8.008, 29.97002997002997, 3.22
+KoNViD_1k_videos/8455309721.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/12020933633.mp4, 8.0, 25.0, 3.36
+KoNViD_1k_videos/6149591212.mp4, 8.008, 29.97002997002997, 2.14
+KoNViD_1k_videos/3710887216.mp4, 8.008, 29.97002997002997, 2.92
+KoNViD_1k_videos/7385692814.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/7341359052.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/6913535960.mp4, 8.0, 25.0, 4.106299213
+KoNViD_1k_videos/3722785424.mp4, 8.008, 29.97002997002997, 3.571428571
+KoNViD_1k_videos/7029270875.mp4, 8.008, 23.976023976023978, 3.94
+KoNViD_1k_videos/5115193359.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/5115335471.mp4, 8.008, 29.97002997002997, 2.16
+KoNViD_1k_videos/6324422954.mp4, 8.008, 29.97002997002997, 2.76
+KoNViD_1k_videos/5029525594.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/9060241382.mp4, 8.0, 25.0, 4.04
+KoNViD_1k_videos/5266893553.mp4, 8.008, 29.97002997002997, 1.74
+KoNViD_1k_videos/6864942042.mp4, 8.0, 24.0, 2.72
+KoNViD_1k_videos/4867899444.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/8005708131.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/7556474634.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/9480678073.mp4, 8.008008, 29.97, 4.04
+KoNViD_1k_videos/8728910927.mp4, 8.0, 24.0, 2.78
+KoNViD_1k_videos/4803433208.mp4, 8.008, 29.97002997002997, 3.823529412
+KoNViD_1k_videos/10013374164.mp4, 8.008008, 29.97, 3.74
+KoNViD_1k_videos/9095814754.mp4, 8.008008, 23.976, 3.8
+KoNViD_1k_videos/5769119862.mp4, 8.008, 29.97002997002997, 3.36
+KoNViD_1k_videos/9536307250.mp4, 8.008008, 29.97, 3.16
+KoNViD_1k_videos/4067736860.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/5901642116.mp4, 8.008, 29.97002997002997, 2.74
+KoNViD_1k_videos/5418971855.mp4, 8.008, 29.97002997002997, 2.68
+KoNViD_1k_videos/6946901845.mp4, 8.008, 29.97002997002997, 3.46
+KoNViD_1k_videos/5475555707.mp4, 8.008, 29.97002997002997, 1.74
+KoNViD_1k_videos/6275674780.mp4, 8.0, 24.0, 1.82
+KoNViD_1k_videos/8631895364.mp4, 8.008, 29.97002997002997, 3.66
+KoNViD_1k_videos/11864752706.mp4, 8.008, 29.97002997002997, 3.9
+KoNViD_1k_videos/6208528782.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/6280406572.mp4, 8.0, 24.0, 2.62
+KoNViD_1k_videos/6469679863.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/5379210063.mp4, 8.008, 29.97002997002997, 2.84
+KoNViD_1k_videos/3469529939.mp4, 8.008, 29.97002997002997, 2.8
+KoNViD_1k_videos/7897815844.mp4, 8.008, 29.97002997002997, 2.96
+KoNViD_1k_videos/6900620920.mp4, 8.0, 24.0, 4.04
+KoNViD_1k_videos/6120978816.mp4, 8.0, 24.0, 2.04
+KoNViD_1k_videos/6340576433.mp4, 8.008, 29.97002997002997, 3.46
+KoNViD_1k_videos/5369095478.mp4, 8.008, 29.97002997002997, 2.6
+KoNViD_1k_videos/5246827285.mp4, 8.0, 24.0, 3.16
+KoNViD_1k_videos/9814654176.mp4, 8.008008, 29.97, 1.94
+KoNViD_1k_videos/7585780552.mp4, 8.008, 29.97002997002997, 3.52
+KoNViD_1k_videos/10925288616.mp4, 8.0, 24.0, 1.46
+KoNViD_1k_videos/10249998364.mp4, 7.799466, 23.976, 3.4
+KoNViD_1k_videos/6300501644.mp4, 8.0, 24.0, 3.4
+KoNViD_1k_videos/5688399126.mp4, 8.008, 29.97002997002997, 3.78
+KoNViD_1k_videos/4472730048.mp4, 8.008, 29.97002997002997, 1.622406639
+KoNViD_1k_videos/12461764004.mp4, 8.008, 23.976023976023978, 2.16
+KoNViD_1k_videos/9812197286.mp4, 8.008008, 29.97, 2.88
+KoNViD_1k_videos/9633321480.mp4, 8.0, 24.0, 4.08
+KoNViD_1k_videos/5633788395.mp4, 8.008, 29.97002997002997, 3.04
+KoNViD_1k_videos/5611452609.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/8355599976.mp4, 8.008, 23.976023976023978, 3.62
+KoNViD_1k_videos/8991258489.mp4, 8.0, 24.0, 3.86
+KoNViD_1k_videos/9633125766.mp4, 8.008008, 29.97, 3.42
+KoNViD_1k_videos/6242642418.mp4, 8.008, 29.97002997002997, 3.398373984
+KoNViD_1k_videos/8304230962.mp4, 8.0, 24.0, 2.32
+KoNViD_1k_videos/5250183455.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/8731237129.mp4, 8.0, 24.0, 2.48
+KoNViD_1k_videos/10016324794.mp4, 8.008008, 23.976, 2.96
+KoNViD_1k_videos/6028497716.mp4, 8.0, 24.0, 1.48
+KoNViD_1k_videos/4996677282.mp4, 8.008, 29.97002997002997, 2.56
+KoNViD_1k_videos/4265470174.mp4, 8.008, 29.97002997002997, 1.56
+KoNViD_1k_videos/10411421606.mp4, 8.008008, 29.97, 2.52
+KoNViD_1k_videos/9195826668.mp4, 8.008008, 29.97, 3.16
+KoNViD_1k_videos/10314276135.mp4, 8.008008, 29.97, 3.22
+KoNViD_1k_videos/5974403907.mp4, 8.008, 23.976023976023978, 3.1
+KoNViD_1k_videos/4837816091.mp4, 8.008, 23.976023976023978, 3.54
+KoNViD_1k_videos/11913590004.mp4, 8.0, 24.0, 2.96
+KoNViD_1k_videos/5489207006.mp4, 8.008, 29.97002997002997, 4.04
+KoNViD_1k_videos/8741017589.mp4, 8.008008, 29.97, 3.676470588
+KoNViD_1k_videos/4211514567.mp4, 8.0, 24.0, 2.74
+KoNViD_1k_videos/3516162818.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/5206708965.mp4, 8.0, 24.0, 3.28
+KoNViD_1k_videos/10115791113.mp4, 8.008008, 29.97, 3.68
+KoNViD_1k_videos/4283455837.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/6184163826.mp4, 8.008, 29.97002997002997, 1.52
+KoNViD_1k_videos/9564154386.mp4, 8.008008, 29.97, 2.1
+KoNViD_1k_videos/8415984918.mp4, 8.008, 23.976023976023978, 2.38
+KoNViD_1k_videos/7616950966.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/11367153336.mp4, 8.008, 29.97002997002997, 2.7
+KoNViD_1k_videos/7396603710.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/8955950667.mp4, 8.008008, 29.97, 3.66
+KoNViD_1k_videos/7066660513.mp4, 8.008, 29.97002997002997, 3.8
+KoNViD_1k_videos/5307981340.mp4, 8.008, 29.97002997002997, 3.52
+KoNViD_1k_videos/8182527128.mp4, 8.008, 29.97002997002997, 3.96
+KoNViD_1k_videos/5329960963.mp4, 8.0, 24.0, 3.04
+KoNViD_1k_videos/5564209652.mp4, 8.008, 29.97002997002997, 2.46
+KoNViD_1k_videos/4088281204.mp4, 8.008, 29.97002997002997, 3.43776824
+KoNViD_1k_videos/3744752110.mp4, 8.0, 24.0, 2.48
+KoNViD_1k_videos/5988574543.mp4, 8.008, 29.97002997002997, 2.9
+KoNViD_1k_videos/8587927758.mp4, 8.0, 24.0, 2.14
+KoNViD_1k_videos/6021943274.mp4, 8.008, 23.976023976023978, 3.14
+KoNViD_1k_videos/4328729913.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/8118947664.mp4, 8.008, 29.97002997002997, 3.76
+KoNViD_1k_videos/7517202292.mp4, 8.008, 29.97002997002997, 2.38
+KoNViD_1k_videos/5266882983.mp4, 8.008, 29.97002997002997, 2.3
+KoNViD_1k_videos/5318369391.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/8729448612.mp4, 8.008008, 29.97, 2.7
+KoNViD_1k_videos/4495254692.mp4, 7.966291999999999, 23.976023976023978, 3.9
+KoNViD_1k_videos/7192436890.mp4, 8.008, 29.97002997002997, 2.78
+KoNViD_1k_videos/6027407703.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/6294328198.mp4, 8.008, 29.97002997002997, 2.52
+KoNViD_1k_videos/5956265529.mp4, 8.008, 29.97002997002997, 2.04
+KoNViD_1k_videos/4783174196.mp4, 8.008, 29.97002997002997, 2.7
+KoNViD_1k_videos/8403451337.mp4, 8.008, 29.97002997002997, 2.78
+KoNViD_1k_videos/6621826551.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/6238991743.mp4, 8.008, 29.97002997002997, 1.68
+KoNViD_1k_videos/9515797648.mp4, 8.008008, 29.97, 3.78
+KoNViD_1k_videos/8152764264.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/5455582932.mp4, 8.008, 29.97002997002997, 3.352272727
+KoNViD_1k_videos/5079146602.mp4, 8.0, 24.0, 3.5
+KoNViD_1k_videos/3975461732.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/4870064087.mp4, 8.008, 29.97002997002997, 2.96
+KoNViD_1k_videos/10838546955.mp4, 8.008008, 29.97, 3.0
+KoNViD_1k_videos/4204485800.mp4, 8.008, 29.97002997002997, 1.72972973
+KoNViD_1k_videos/3431305644.mp4, 8.0, 24.0, 3.06
+KoNViD_1k_videos/9830655414.mp4, 8.008008, 29.97, 3.56
+KoNViD_1k_videos/10019485755.mp4, 8.008008, 29.97, 3.7
+KoNViD_1k_videos/5279319629.mp4, 8.008, 29.97002997002997, 3.88
+KoNViD_1k_videos/5363205423.mp4, 8.0, 24.0, 3.36
+KoNViD_1k_videos/6497857295.mp4, 8.008, 29.97002997002997, 2.38
+KoNViD_1k_videos/12893008605.mp4, 8.008008, 29.97, 3.55465587
+KoNViD_1k_videos/6924551254.mp4, 8.008, 29.97002997002997, 2.72
+KoNViD_1k_videos/4281880002.mp4, 8.0, 24.0, 3.2
+KoNViD_1k_videos/5295005655.mp4, 8.0, 24.0, 2.72
+KoNViD_1k_videos/5506184490.mp4, 8.008, 29.97002997002997, 2.04
+KoNViD_1k_videos/10340094005.mp4, 8.008008, 29.97, 3.32
+KoNViD_1k_videos/12788630734.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/4279268661.mp4, 8.008, 29.97002997002997, 3.86
+KoNViD_1k_videos/6279406877.mp4, 8.008, 29.97002997002997, 3.5
+KoNViD_1k_videos/8253260683.mp4, 8.0, 24.0, 3.08
+KoNViD_1k_videos/10625504535.mp4, 8.008008, 29.97, 2.1
+KoNViD_1k_videos/5754220740.mp4, 8.0, 24.0, 2.74
+KoNViD_1k_videos/5360605734.mp4, 8.008, 29.97002997002997, 3.84
+KoNViD_1k_videos/10931366544.mp4, 8.008, 23.976023976023978, 3.64
+KoNViD_1k_videos/7520968610.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/8459254855.mp4, 8.008, 29.97002997002997, 2.74
+KoNViD_1k_videos/9369433766.mp4, 7.5492159999999995, 23.976, 1.52
+KoNViD_1k_videos/6838572151.mp4, 8.0, 24.0, 2.6
+KoNViD_1k_videos/5419510404.mp4, 8.0, 25.0, 2.26
+KoNViD_1k_videos/8685321568.mp4, 7.5416669999999995, 24.0, 2.18
+KoNViD_1k_videos/8281880253.mp4, 8.0, 24.0, 2.4
+KoNViD_1k_videos/7358630018.mp4, 8.008, 23.976023976023978, 3.6
+KoNViD_1k_videos/7622034108.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/8643275693.mp4, 8.008, 29.97002997002997, 3.22
+KoNViD_1k_videos/3263720175.mp4, 8.008, 23.976023976023978, 3.04
+KoNViD_1k_videos/6223028206.mp4, 8.008, 29.97002997002997, 1.68
+KoNViD_1k_videos/8536919744.mp4, 8.0, 24.0, 1.22
+KoNViD_1k_videos/6291796040.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/12032110574.mp4, 8.0, 25.0, 2.64
+KoNViD_1k_videos/7481217984.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/8052505414.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/9908050493.mp4, 7.540874, 29.97, 2.96
+KoNViD_1k_videos/5409101281.mp4, 8.008, 29.97002997002997, 2.54
+KoNViD_1k_videos/7812985864.mp4, 8.008, 29.97002997002997, 2.76
+KoNViD_1k_videos/11086166076.mp4, 8.008008, 29.97, 3.28
+KoNViD_1k_videos/4823416929.mp4, 8.008, 29.97002997002997, 1.98
+KoNViD_1k_videos/6408210773.mp4, 8.008, 23.976023976023978, 2.92
+KoNViD_1k_videos/5678628993.mp4, 8.008, 29.97002997002997, 3.567226891
+KoNViD_1k_videos/5777975441.mp4, 8.008, 29.97002997002997, 2.3
+KoNViD_1k_videos/5392736832.mp4, 8.0, 24.0, 2.42
+KoNViD_1k_videos/12279842993.mp4, 8.008008, 29.97, 3.72
+KoNViD_1k_videos/3629928686.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/8717841410.mp4, 8.008, 29.97002997002997, 3.22
+KoNViD_1k_videos/5660712402.mp4, 8.008, 29.97002997002997, 3.82
+KoNViD_1k_videos/3825035703.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/8965205906.mp4, 8.008008, 29.97, 3.6
+KoNViD_1k_videos/8156543755.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/7541374876.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/7938991482.mp4, 8.008, 29.97002997002997, 2.52
+KoNViD_1k_videos/4905336825.mp4, 8.008, 29.97002997002997, 1.828125
+KoNViD_1k_videos/8921577578.mp4, 8.008008, 29.97, 3.48
+KoNViD_1k_videos/6518466029.mp4, 8.0, 24.0, 3.02
+KoNViD_1k_videos/6677269327.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/7453163014.mp4, 8.0, 24.0, 3.02
+KoNViD_1k_videos/5705527370.mp4, 8.008, 29.97002997002997, 3.74
+KoNViD_1k_videos/9464591256.mp4, 8.008008, 23.976, 3.98
+KoNViD_1k_videos/11013388203.mp4, 8.008, 23.976023976023978, 3.76
+KoNViD_1k_videos/6021794037.mp4, 8.008, 29.97002997002997, 3.02
+KoNViD_1k_videos/4933934170.mp4, 8.008, 29.97002997002997, 1.96
+KoNViD_1k_videos/4761429801.mp4, 8.0, 24.0, 3.08
+KoNViD_1k_videos/8424428827.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/10008004183.mp4, 8.0, 25.0, 4.06
+KoNViD_1k_videos/11383768246.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/9083519636.mp4, 8.008008, 29.97, 2.86
+KoNViD_1k_videos/4932309592.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/5882410880.mp4, 8.008, 29.97002997002997, 3.52
+KoNViD_1k_videos/6296962313.mp4, 8.008, 29.97002997002997, 2.64
+KoNViD_1k_videos/7992241767.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/6760172457.mp4, 8.0, 24.0, 3.24
+KoNViD_1k_videos/5459565205.mp4, 7.9079, 29.97002997002997, 3.52
+KoNViD_1k_videos/5260933475.mp4, 8.0, 24.0, 3.08
+KoNViD_1k_videos/5147152978.mp4, 8.008, 29.97002997002997, 2.597609562
+KoNViD_1k_videos/8751856572.mp4, 8.008008, 29.97, 3.9
+KoNViD_1k_videos/7977126554.mp4, 8.008, 29.97002997002997, 3.96
+KoNViD_1k_videos/12633847193.mp4, 8.008008, 29.97, 2.92
+KoNViD_1k_videos/12704590184.mp4, 8.008, 23.976023976023978, 3.9
+KoNViD_1k_videos/8195185860.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/4526948676.mp4, 8.008, 29.97002997002997, 3.34
+KoNViD_1k_videos/8953002659.mp4, 8.008008, 29.97, 3.9
+KoNViD_1k_videos/9478631712.mp4, 8.008008, 29.97, 3.706349206
+KoNViD_1k_videos/9382662040.mp4, 8.0, 24.0, 2.76
+KoNViD_1k_videos/3763235915.mp4, 8.0, 24.0, 2.343873518
+KoNViD_1k_videos/9617144245.mp4, 7.741073999999999, 29.97, 3.82
+KoNViD_1k_videos/9917916796.mp4, 8.008008, 29.97, 4.04
+KoNViD_1k_videos/10595581344.mp4, 8.008008, 29.97, 4.14
+KoNViD_1k_videos/5479646341.mp4, 8.008, 29.97002997002997, 3.02
+KoNViD_1k_videos/8331470149.mp4, 8.008, 23.976023976023978, 3.02
+KoNViD_1k_videos/6320892865.mp4, 8.008, 23.976023976023978, 2.84
+KoNViD_1k_videos/7287123594.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/9659398278.mp4, 8.008008, 29.97, 3.3
+KoNViD_1k_videos/5150451769.mp4, 8.0, 25.0, 1.84
+KoNViD_1k_videos/8758952624.mp4, 8.008008, 29.97, 4.3
+KoNViD_1k_videos/5653702857.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/8536800170.mp4, 8.0, 24.0, 3.12
+KoNViD_1k_videos/7591242800.mp4, 8.0, 25.0, 3.16
+KoNViD_1k_videos/10102107193.mp4, 8.0, 25.0, 2.18
+KoNViD_1k_videos/4177089041.mp4, 8.008, 29.97002997002997, 1.66
+KoNViD_1k_videos/8336226077.mp4, 8.008, 29.97002997002997, 2.08
+KoNViD_1k_videos/6333567508.mp4, 8.008, 29.97002997002997, 3.18
+KoNViD_1k_videos/3591184129.mp4, 8.008, 29.97002997002997, 2.58
+KoNViD_1k_videos/8195899669.mp4, 8.008, 29.97002997002997, 2.06
+KoNViD_1k_videos/5101801222.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/8039435442.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/6629527951.mp4, 8.0, 24.0, 2.25498008
+KoNViD_1k_videos/8610122536.mp4, 8.008, 29.97002997002997, 3.58
+KoNViD_1k_videos/4773686738.mp4, 8.008, 29.97002997002997, 2.32
+KoNViD_1k_videos/6661657005.mp4, 8.008, 29.97002997002997, 1.9
+KoNViD_1k_videos/9726323401.mp4, 8.008008, 29.97, 3.272
+KoNViD_1k_videos/6273360207.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/7057951213.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/6849313907.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/8663100289.mp4, 8.0, 24.0, 2.034883721
+KoNViD_1k_videos/8583146942.mp4, 8.008, 29.97002997002997, 2.54
+KoNViD_1k_videos/9651230522.mp4, 8.008008, 29.97, 2.56
+KoNViD_1k_videos/8918155991.mp4, 8.008008, 29.97, 3.46
+KoNViD_1k_videos/8587203405.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/5941263615.mp4, 8.0, 25.0, 3.54
+KoNViD_1k_videos/4850105788.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/12099271596.mp4, 8.0, 24.0, 2.4
+KoNViD_1k_videos/6625655725.mp4, 8.008, 29.97002997002997, 3.62
+KoNViD_1k_videos/5133219451.mp4, 7.708333, 24.0, 2.32
+KoNViD_1k_videos/7734422826.mp4, 8.008, 29.97002997002997, 2.82
+KoNViD_1k_videos/5134417428.mp4, 8.008, 29.97002997002997, 2.26
+KoNViD_1k_videos/8487362151.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/4956521693.mp4, 8.008, 29.97002997002997, 2.8
+KoNViD_1k_videos/9725289895.mp4, 8.008008, 29.97, 3.58
+KoNViD_1k_videos/6157684615.mp4, 8.008, 29.97002997002997, 3.36
+KoNViD_1k_videos/8527434285.mp4, 8.008, 29.97002997002997, 2.02
+KoNViD_1k_videos/3583315023.mp4, 8.0, 24.0, 2.84
+KoNViD_1k_videos/4801224837.mp4, 8.0, 24.0, 3.46
+KoNViD_1k_videos/5739175520.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/4775312025.mp4, 8.008, 29.97002997002997, 2.08
+KoNViD_1k_videos/12186470506.mp4, 8.008, 29.97002997002997, 2.9
+KoNViD_1k_videos/8616015387.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/6096558457.mp4, 8.008, 29.97002997002997, 2.6
+KoNViD_1k_videos/8331282929.mp4, 8.008, 23.976023976023978, 3.76
+KoNViD_1k_videos/4733600732.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/6381011947.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/5170915466.mp4, 8.008, 29.97002997002997, 2.88
+KoNViD_1k_videos/7644822866.mp4, 8.008, 29.97002997002997, 3.78
+KoNViD_1k_videos/8604343265.mp4, 8.008, 23.976023976023978, 2.54
+KoNViD_1k_videos/5273433233.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/3627681183.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/5815295661.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/6015938306.mp4, 8.008, 29.97002997002997, 2.5
+KoNViD_1k_videos/7848940332.mp4, 8.008, 29.97002997002997, 1.86
+KoNViD_1k_videos/12330492194.mp4, 8.008008, 29.97, 3.26
+KoNViD_1k_videos/5914289359.mp4, 8.008, 29.97002997002997, 3.56
+KoNViD_1k_videos/3861648412.mp4, 8.0, 24.0, 2.68
+KoNViD_1k_videos/5553355582.mp4, 8.008, 29.97002997002997, 2.22
+KoNViD_1k_videos/6210381577.mp4, 8.0, 24.0, 3.22
+KoNViD_1k_videos/11469243925.mp4, 8.008008, 29.97, 2.28
+KoNViD_1k_videos/11944451643.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/5014400375.mp4, 8.008, 29.97002997002997, 2.28
+KoNViD_1k_videos/8724380666.mp4, 8.008, 29.97002997002997, 1.4
+KoNViD_1k_videos/8541047953.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/11095294653.mp4, 8.008008, 29.97, 1.86
+KoNViD_1k_videos/8886176135.mp4, 8.008008, 29.97, 3.56
+KoNViD_1k_videos/4969714039.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/13255378043.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/11617643563.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/5999732346.mp4, 8.0, 24.0, 2.9
+KoNViD_1k_videos/10837320963.mp4, 8.0, 25.0, 3.7
+KoNViD_1k_videos/8129160337.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/6794539977.mp4, 8.0, 25.0, 3.4
+KoNViD_1k_videos/8733102851.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/4373113207.mp4, 8.008, 29.97002997002997, 3.52
+KoNViD_1k_videos/6340146486.mp4, 8.0, 25.0, 2.48
+KoNViD_1k_videos/4652600104.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/5559135240.mp4, 8.0, 24.0, 2.94
+KoNViD_1k_videos/6985991227.mp4, 8.008, 29.97002997002997, 1.62
+KoNViD_1k_videos/5632758635.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/3339962845.mp4, 8.008, 29.97002997002997, 4.64
+KoNViD_1k_videos/5636445745.mp4, 8.008, 29.97002997002997, 2.52
+KoNViD_1k_videos/3770747002.mp4, 8.008, 29.97002997002997, 1.86
+KoNViD_1k_videos/8643336872.mp4, 8.008, 29.97002997002997, 1.68
+KoNViD_1k_videos/8588854353.mp4, 8.008, 23.976023976023978, 3.88
+KoNViD_1k_videos/5668795950.mp4, 8.0, 24.0, 2.86
+KoNViD_1k_videos/6497033987.mp4, 8.008, 29.97002997002997, 1.8
+KoNViD_1k_videos/4677538291.mp4, 8.008, 29.97002997002997, 1.96
+KoNViD_1k_videos/12274927193.mp4, 8.008, 23.976023976023978, 2.98
+KoNViD_1k_videos/9256674744.mp4, 8.008008, 29.97, 2.26
+KoNViD_1k_videos/5040855662.mp4, 8.008, 29.97002997002997, 3.18
+KoNViD_1k_videos/9093682547.mp4, 8.008008, 23.976, 3.1
+KoNViD_1k_videos/8716622241.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/5935222726.mp4, 8.008, 29.97002997002997, 1.86
+KoNViD_1k_videos/10149337843.mp4, 8.008008, 29.97, 3.94
+KoNViD_1k_videos/5012335397.mp4, 8.008, 29.97002997002997, 2.74
+KoNViD_1k_videos/3489123822.mp4, 8.0, 24.0, 3.8
+KoNViD_1k_videos/6408909511.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/7986004475.mp4, 8.0, 25.0, 3.58
+KoNViD_1k_videos/5028474013.mp4, 8.008, 29.97002997002997, 3.52
+KoNViD_1k_videos/5110540070.mp4, 8.008, 29.97002997002997, 2.64
+KoNViD_1k_videos/5067647428.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/8178490084.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/5589046662.mp4, 8.008, 29.97002997002997, 3.134453782
+KoNViD_1k_videos/5541058845.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/5340069287.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/6528988807.mp4, 8.0, 24.0, 3.94488189
+KoNViD_1k_videos/9586728257.mp4, 8.008008, 29.97, 3.56
+KoNViD_1k_videos/12836715003.mp4, 8.008008, 29.97, 3.3
+KoNViD_1k_videos/11803329033.mp4, 8.008008, 29.97, 3.56
+KoNViD_1k_videos/10417424703.mp4, 8.008008, 29.97, 1.88
+KoNViD_1k_videos/8147563408.mp4, 7.640967, 29.97002997002997, 3.58
+KoNViD_1k_videos/6293871635.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/9047671138.mp4, 8.008008, 29.97, 1.98
+KoNViD_1k_videos/4406917423.mp4, 8.008, 23.976023976023978, 3.18
+KoNViD_1k_videos/4367069947.mp4, 8.0, 24.0, 2.88
+KoNViD_1k_videos/5020714258.mp4, 8.008, 23.976023976023978, 2.4
+KoNViD_1k_videos/4744073127.mp4, 8.008, 29.97002997002997, 1.42
+KoNViD_1k_videos/4843048363.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/7734563104.mp4, 8.008, 23.976023976023978, 2.7
+KoNViD_1k_videos/8670600555.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/6582418637.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/7383232724.mp4, 8.008, 29.97002997002997, 2.6
+KoNViD_1k_videos/6343817087.mp4, 8.008, 29.97002997002997, 2.64
+KoNViD_1k_videos/5827345790.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/4922109288.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/6932589551.mp4, 8.008, 29.97002997002997, 3.62
+KoNViD_1k_videos/6338210961.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/3437816776.mp4, 8.0, 24.0, 3.5
+KoNViD_1k_videos/8431144343.mp4, 8.0, 24.0, 3.82
+KoNViD_1k_videos/5893119747.mp4, 8.0, 25.0, 4.12
+KoNViD_1k_videos/4270576961.mp4, 8.008, 29.97002997002997, 2.54
+KoNViD_1k_videos/7633708420.mp4, 8.008, 23.976023976023978, 2.78
+KoNViD_1k_videos/4383007820.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/8582023917.mp4, 8.008, 29.97002997002997, 2.04
+KoNViD_1k_videos/5504464599.mp4, 8.0, 24.0, 1.94
+KoNViD_1k_videos/8532196494.mp4, 8.008, 29.97002997002997, 2.46
+KoNViD_1k_videos/8201673711.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/6879288904.mp4, 8.0, 24.0, 2.78
+KoNViD_1k_videos/9513903384.mp4, 8.008008, 29.97, 4.04
+KoNViD_1k_videos/4033322131.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/10403478556.mp4, 8.008008, 23.976, 3.68
+KoNViD_1k_videos/8207372644.mp4, 8.008, 29.97002997002997, 3.24
+KoNViD_1k_videos/6131895268.mp4, 8.008, 29.97002997002997, 1.94
+KoNViD_1k_videos/7703064692.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/7500126174.mp4, 8.0, 24.0, 3.6
+KoNViD_1k_videos/6268574586.mp4, 8.008, 23.976023976023978, 1.8
+KoNViD_1k_videos/6373417661.mp4, 8.008, 23.976023976023978, 3.72
+KoNViD_1k_videos/5693366090.mp4, 8.008, 29.97002997002997, 3.62
+KoNViD_1k_videos/3358799122.mp4, 8.008, 29.97002997002997, 3.785714286
+KoNViD_1k_videos/9812351333.mp4, 8.008008, 29.97, 3.74
+KoNViD_1k_videos/10657154643.mp4, 8.008008, 29.97, 3.58
+KoNViD_1k_videos/4184711791.mp4, 8.0, 24.0, 3.28
+KoNViD_1k_videos/8552723245.mp4, 8.008, 23.976023976023978, 3.58
+KoNViD_1k_videos/8842319020.mp4, 8.008008, 23.976, 2.86
+KoNViD_1k_videos/7760313026.mp4, 8.008, 29.97002997002997, 3.36
+KoNViD_1k_videos/8069880437.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/12276570315.mp4, 8.0, 24.0, 3.365217391
+KoNViD_1k_videos/5900513019.mp4, 8.008, 29.97002997002997, 3.231660232
+KoNViD_1k_videos/6696625151.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/6792594754.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/10053703034.mp4, 8.008008, 29.97, 3.46
+KoNViD_1k_videos/5621520605.mp4, 8.008, 29.97002997002997, 3.34
+KoNViD_1k_videos/8144762018.mp4, 8.0, 24.0, 2.1
+KoNViD_1k_videos/6095633507.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/6299595444.mp4, 8.008, 29.97002997002997, 1.96
+KoNViD_1k_videos/7681404078.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/4484771725.mp4, 8.008, 29.97002997002997, 3.62
+KoNViD_1k_videos/6722692103.mp4, 8.0, 24.0, 3.5
+KoNViD_1k_videos/10201521555.mp4, 8.008008, 23.976, 2.3
+KoNViD_1k_videos/3218223446.mp4, 8.008, 29.97002997002997, 2.92
+KoNViD_1k_videos/5880475570.mp4, 8.0, 25.0, 2.695652174
+KoNViD_1k_videos/6471207533.mp4, 8.0, 24.0, 3.32
+KoNViD_1k_videos/7425861358.mp4, 8.008, 29.97002997002997, 3.46
+KoNViD_1k_videos/11359714186.mp4, 8.008008, 29.97, 2.4
+KoNViD_1k_videos/3546788918.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/9680817388.mp4, 8.008008, 29.97, 3.84
+KoNViD_1k_videos/8699104416.mp4, 8.008, 23.976023976023978, 2.48
+KoNViD_1k_videos/5672502803.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/7542896986.mp4, 8.0, 24.0, 3.02
+KoNViD_1k_videos/8235549337.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/4675931801.mp4, 8.008, 29.97002997002997, 4.16
+KoNViD_1k_videos/5487345956.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/8252040986.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/6079875162.mp4, 8.008, 29.97002997002997, 1.6
+KoNViD_1k_videos/5927908371.mp4, 8.008, 29.97002997002997, 1.6
+KoNViD_1k_videos/5048882871.mp4, 8.0, 24.0, 1.84
+KoNViD_1k_videos/8772282000.mp4, 8.008, 29.97002997002997, 3.24
+KoNViD_1k_videos/3904711682.mp4, 8.008, 29.97002997002997, 3.04
+KoNViD_1k_videos/5119193300.mp4, 8.008, 29.97002997002997, 2.46
+KoNViD_1k_videos/5107749210.mp4, 8.008, 29.97002997002997, 2.36
+KoNViD_1k_videos/9298912586.mp4, 8.0, 24.0, 2.5
+KoNViD_1k_videos/9234105504.mp4, 8.008008, 23.976, 2.26
+KoNViD_1k_videos/6323632937.mp4, 8.0, 24.0, 1.82
+KoNViD_1k_videos/8320880668.mp4, 8.0, 24.0, 2.22
+KoNViD_1k_videos/5561000398.mp4, 8.0, 24.0, 2.88
+KoNViD_1k_videos/8731159058.mp4, 8.008008, 29.97, 2.2
+KoNViD_1k_videos/6058824138.mp4, 8.008, 29.97002997002997, 2.74
+KoNViD_1k_videos/5214650735.mp4, 8.008, 29.97002997002997, 3.18
+KoNViD_1k_videos/5204063512.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/6346933640.mp4, 8.008, 23.976023976023978, 2.6
+KoNViD_1k_videos/4897814885.mp4, 8.0, 25.0, 2.46
+KoNViD_1k_videos/6527804569.mp4, 8.0, 24.0, 3.12
+KoNViD_1k_videos/4834450895.mp4, 8.008, 29.97002997002997, 1.74
+KoNViD_1k_videos/3339387387.mp4, 8.008, 29.97002997002997, 2.42
+KoNViD_1k_videos/4942462682.mp4, 8.008, 29.97002997002997, 3.5
+KoNViD_1k_videos/8299847304.mp4, 8.0, 24.0, 2.96
+KoNViD_1k_videos/4801560427.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/7849542524.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/12989512873.mp4, 8.008008, 23.976, 3.24
+KoNViD_1k_videos/8467238180.mp4, 8.008, 29.97002997002997, 3.18
+KoNViD_1k_videos/8222762296.mp4, 8.008, 29.97002997002997, 1.66
+KoNViD_1k_videos/12899309364.mp4, 8.008008, 29.97, 3.1
+KoNViD_1k_videos/6226852017.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/8331955206.mp4, 8.008, 29.97002997002997, 3.84
+KoNViD_1k_videos/3488252741.mp4, 8.008, 23.976023976023978, 3.54
+KoNViD_1k_videos/5972627544.mp4, 8.0, 24.0, 3.04
+KoNViD_1k_videos/12255205254.mp4, 8.0, 25.0, 3.96
+KoNViD_1k_videos/4629711727.mp4, 8.0, 24.0, 2.3
+KoNViD_1k_videos/6268902756.mp4, 8.008, 23.976023976023978, 1.843137255
+KoNViD_1k_videos/4889669313.mp4, 8.008, 29.97002997002997, 2.52
+KoNViD_1k_videos/7556185878.mp4, 8.0, 24.0, 2.82
+KoNViD_1k_videos/8995404511.mp4, 8.008008, 29.97, 3.798245614
+KoNViD_1k_videos/5941894437.mp4, 8.008, 29.97002997002997, 2.48
+KoNViD_1k_videos/4414516665.mp4, 8.0, 25.0, 3.02
+KoNViD_1k_videos/5879908897.mp4, 8.0, 25.0, 3.22
+KoNViD_1k_videos/5130906896.mp4, 8.0, 24.0, 3.38
+KoNViD_1k_videos/6059351914.mp4, 8.0, 24.0, 2.24
+KoNViD_1k_videos/8751538577.mp4, 8.008008, 23.976, 4.4
+KoNViD_1k_videos/4363644898.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/6995540661.mp4, 8.008, 29.97002997002997, 3.18
+KoNViD_1k_videos/10804894696.mp4, 8.008, 23.976023976023978, 3.12
+KoNViD_1k_videos/10408715786.mp4, 8.008008, 29.97, 3.253731343
+KoNViD_1k_videos/7136906983.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/8366966088.mp4, 8.008, 29.97002997002997, 2.84
+KoNViD_1k_videos/6849945228.mp4, 8.008, 23.976023976023978, 2.91796875
+KoNViD_1k_videos/7548330816.mp4, 8.008, 23.976023976023978, 3.04
+KoNViD_1k_videos/8479579728.mp4, 8.008, 29.97002997002997, 2.28
+KoNViD_1k_videos/6716265511.mp4, 8.0, 24.0, 3.54
+KoNViD_1k_videos/5355837503.mp4, 8.008, 23.976023976023978, 4.1
+KoNViD_1k_videos/10650618913.mp4, 8.008008, 29.97, 1.7
+KoNViD_1k_videos/7917797604.mp4, 8.008, 29.97002997002997, 2.105660377
+KoNViD_1k_videos/7532463480.mp4, 8.008, 29.97002997002997, 2.792828685
+KoNViD_1k_videos/9542585542.mp4, 8.008008, 29.97, 3.82
+KoNViD_1k_videos/8429349812.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/5996750151.mp4, 8.0, 24.0, 3.0
+KoNViD_1k_videos/5865474554.mp4, 8.008, 29.97002997002997, 2.38
+KoNViD_1k_videos/5245445672.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/4876943348.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/9434801897.mp4, 8.0, 25.0, 3.34
+KoNViD_1k_videos/3963620197.mp4, 8.008, 23.976023976023978, 4.02
+KoNViD_1k_videos/8976267501.mp4, 8.008008, 29.97, 3.0
+KoNViD_1k_videos/7395710342.mp4, 8.008, 29.97002997002997, 2.92
+KoNViD_1k_videos/7208119980.mp4, 8.008, 29.97002997002997, 2.93385214
+KoNViD_1k_videos/12893918213.mp4, 8.008008, 29.97, 3.599236641
+KoNViD_1k_videos/8675766145.mp4, 8.008, 29.97002997002997, 2.5
+KoNViD_1k_videos/13154058393.mp4, 8.008, 23.976023976023978, 3.64
+KoNViD_1k_videos/5082292308.mp4, 8.008, 29.97002997002997, 3.64
+KoNViD_1k_videos/3588022983.mp4, 8.0, 24.0, 3.66
+KoNViD_1k_videos/5458000924.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/7141804343.mp4, 8.008, 29.97002997002997, 2.9
+KoNViD_1k_videos/8078333220.mp4, 8.008, 29.97002997002997, 2.6
+KoNViD_1k_videos/4736973814.mp4, 8.008, 29.97002997002997, 3.46
+KoNViD_1k_videos/6971497024.mp4, 8.008, 29.97002997002997, 3.02
+KoNViD_1k_videos/6369679907.mp4, 8.008, 29.97002997002997, 1.72
+KoNViD_1k_videos/3699539878.mp4, 8.0, 24.0, 3.14
+KoNViD_1k_videos/7983363078.mp4, 8.0, 24.0, 1.9
+KoNViD_1k_videos/9390723258.mp4, 8.008008, 29.97, 3.8
+KoNViD_1k_videos/8172276744.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/10178280003.mp4, 8.008008, 29.97, 1.7
+KoNViD_1k_videos/10716551125.mp4, 8.0, 25.0, 2.22
+KoNViD_1k_videos/4258186896.mp4, 8.008, 29.97002997002997, 2.34
+KoNViD_1k_videos/5078125268.mp4, 8.008, 29.97002997002997, 2.14
+KoNViD_1k_videos/6191735552.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/3521396571.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/8729280042.mp4, 8.0, 24.0, 2.26
+KoNViD_1k_videos/3179649855.mp4, 8.008, 23.976023976023978, 2.88
+KoNViD_1k_videos/8476574247.mp4, 8.008, 29.97002997002997, 3.58
+KoNViD_1k_videos/3819966572.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/6514514177.mp4, 8.008, 23.976023976023978, 3.68
+KoNViD_1k_videos/8229018279.mp4, 8.008, 29.97002997002997, 3.56
+KoNViD_1k_videos/5010279097.mp4, 8.0, 24.0, 3.268907563
+KoNViD_1k_videos/5914119796.mp4, 8.008, 29.97002997002997, 2.44
+KoNViD_1k_videos/11061444876.mp4, 8.008, 29.97002997002997, 2.16
+KoNViD_1k_videos/6304997355.mp4, 8.008, 29.97002997002997, 3.34
+KoNViD_1k_videos/6775766604.mp4, 8.008, 29.97002997002997, 2.96
+KoNViD_1k_videos/5066571585.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/6554989029.mp4, 8.0, 24.0, 2.68
+KoNViD_1k_videos/9058636581.mp4, 8.008008, 29.97, 1.82
+KoNViD_1k_videos/6772996477.mp4, 8.008, 23.976023976023978, 3.72
+KoNViD_1k_videos/5469530975.mp4, 8.008, 29.97002997002997, 2.9
+KoNViD_1k_videos/5162398156.mp4, 8.008, 29.97002997002997, 2.68
+KoNViD_1k_videos/5550087163.mp4, 8.008, 29.97002997002997, 2.46
+KoNViD_1k_videos/5341444640.mp4, 8.008, 23.976023976023978, 3.68
+KoNViD_1k_videos/3590551873.mp4, 8.008, 29.97002997002997, 2.7
+KoNViD_1k_videos/3699089579.mp4, 8.0, 24.0, 3.38
+KoNViD_1k_videos/5218123691.mp4, 8.0, 24.0, 3.04
+KoNViD_1k_videos/4273096444.mp4, 8.008, 29.97002997002997, 3.98
+KoNViD_1k_videos/5733977738.mp4, 8.0, 24.0, 2.06
+KoNViD_1k_videos/6694501209.mp4, 8.008, 29.97002997002997, 2.92
+KoNViD_1k_videos/4227960041.mp4, 8.008, 29.97002997002997, 3.66
+KoNViD_1k_videos/6109922367.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/7360639566.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/7361913318.mp4, 8.008, 23.976023976023978, 2.74
+KoNViD_1k_videos/4763574191.mp4, 8.008, 29.97002997002997, 3.469230769
+KoNViD_1k_videos/5319413177.mp4, 8.0, 24.0, 2.48
+KoNViD_1k_videos/9717753449.mp4, 8.008008, 23.976, 2.36
+KoNViD_1k_videos/5690367902.mp4, 8.008, 29.97002997002997, 3.56
+KoNViD_1k_videos/6137257339.mp4, 8.008, 29.97002997002997, 4.02
+KoNViD_1k_videos/8986658663.mp4, 8.008008, 29.97, 3.2
+KoNViD_1k_videos/12382276334.mp4, 8.0, 25.0, 3.586872587
+KoNViD_1k_videos/4286406568.mp4, 8.008, 29.97002997002997, 2.96
+KoNViD_1k_videos/6798366426.mp4, 8.008, 23.976023976023978, 3.84
+KoNViD_1k_videos/10838718774.mp4, 8.008008, 29.97, 3.04
+KoNViD_1k_videos/7969995368.mp4, 8.008, 29.97002997002997, 3.04
+KoNViD_1k_videos/8967888670.mp4, 8.0, 24.0, 3.32
+KoNViD_1k_videos/5212573386.mp4, 8.008, 29.97002997002997, 3.74
+KoNViD_1k_videos/8440192357.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/7177427229.mp4, 8.008, 23.976023976023978, 2.48
+KoNViD_1k_videos/4417558772.mp4, 8.008, 29.97002997002997, 3.9
+KoNViD_1k_videos/8007634384.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/5993312864.mp4, 8.008, 29.97002997002997, 1.72
+KoNViD_1k_videos/6959318659.mp4, 8.008, 29.97002997002997, 3.04
+KoNViD_1k_videos/9675032717.mp4, 8.0, 25.0, 3.26
+KoNViD_1k_videos/3496871384.mp4, 8.008, 29.97002997002997, 2.08
+KoNViD_1k_videos/6270222640.mp4, 8.008, 29.97002997002997, 2.32
+KoNViD_1k_videos/8414555471.mp4, 8.0, 25.0, 3.9
+KoNViD_1k_videos/5011231037.mp4, 8.008, 29.97002997002997, 3.62195122
+KoNViD_1k_videos/6638842709.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/8646141367.mp4, 8.0, 25.0, 3.8
+KoNViD_1k_videos/4265466447.mp4, 8.0, 24.0, 3.32
+KoNViD_1k_videos/6867937233.mp4, 8.008, 29.97002997002997, 2.72
+KoNViD_1k_videos/9085955167.mp4, 8.008008, 29.97, 3.2
+KoNViD_1k_videos/4549036485.mp4, 8.008, 29.97002997002997, 3.992481203
+KoNViD_1k_videos/6329730783.mp4, 8.0, 24.0, 2.28
+KoNViD_1k_videos/4743183444.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/8667333526.mp4, 8.0, 25.0, 2.62
+KoNViD_1k_videos/6740483069.mp4, 8.008, 29.97002997002997, 2.2421875
+KoNViD_1k_videos/9445782126.mp4, 8.008008, 29.97, 1.52
+KoNViD_1k_videos/6858182534.mp4, 8.008, 29.97002997002997, 2.98
+KoNViD_1k_videos/6135877141.mp4, 8.0, 24.0, 2.82
+KoNViD_1k_videos/6949440526.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/8531461869.mp4, 8.0, 24.0, 2.72
+KoNViD_1k_videos/4751595925.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/8031519344.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/8433382995.mp4, 8.008, 23.976023976023978, 2.22
+KoNViD_1k_videos/4337949901.mp4, 8.0, 24.0, 3.34
+KoNViD_1k_videos/7087189849.mp4, 8.0, 24.0, 1.98
+KoNViD_1k_videos/4313579496.mp4, 8.0, 24.0, 3.76
+KoNViD_1k_videos/8251526312.mp4, 8.008, 29.97002997002997, 2.28
+KoNViD_1k_videos/5819852885.mp4, 8.008, 29.97002997002997, 2.16
+KoNViD_1k_videos/13196703974.mp4, 8.0, 25.0, 3.48
+KoNViD_1k_videos/4281436885.mp4, 8.0, 24.0, 3.494339623
+KoNViD_1k_videos/3810054574.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/11535683914.mp4, 8.008, 29.97002997002997, 3.94
+KoNViD_1k_videos/6598948831.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/8319369964.mp4, 8.008, 29.97002997002997, 2.64
+KoNViD_1k_videos/12387439244.mp4, 8.008008, 29.97, 2.34
+KoNViD_1k_videos/12031712185.mp4, 8.0, 25.0, 2.58
+KoNViD_1k_videos/4506003922.mp4, 8.008, 23.976023976023978, 2.36
+KoNViD_1k_videos/5134310379.mp4, 8.008, 29.97002997002997, 2.12
+KoNViD_1k_videos/5723108521.mp4, 8.008, 29.97002997002997, 3.18
+KoNViD_1k_videos/8322973249.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/6867862287.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/6669111911.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/10244479353.mp4, 8.008008, 29.97, 4.14
+KoNViD_1k_videos/9571377943.mp4, 8.008008, 23.976, 4.28
+KoNViD_1k_videos/5831807090.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/6947112934.mp4, 8.0, 25.0, 4.18
+KoNViD_1k_videos/4297535685.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/5618829039.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/11688667226.mp4, 8.008008, 29.97, 1.94
+KoNViD_1k_videos/8298917863.mp4, 8.0, 24.0, 3.28
+KoNViD_1k_videos/12101500915.mp4, 8.008, 29.97002997002997, 3.88
+KoNViD_1k_videos/9098899158.mp4, 8.008008, 23.976, 3.02
+KoNViD_1k_videos/5391431291.mp4, 8.008, 23.976023976023978, 4.151851852
+KoNViD_1k_videos/8056761129.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/4734666961.mp4, 8.008, 29.97002997002997, 2.82
+KoNViD_1k_videos/9425570317.mp4, 8.0, 24.0, 3.0
+KoNViD_1k_videos/3374751086.mp4, 8.008, 23.976023976023978, 2.24
+KoNViD_1k_videos/9277742312.mp4, 8.008008, 29.97, 3.4
+KoNViD_1k_videos/5516081814.mp4, 8.0, 24.0, 3.6
+KoNViD_1k_videos/12823937005.mp4, 8.008008, 29.97, 3.4
+KoNViD_1k_videos/5208145791.mp4, 8.008, 29.97002997002997, 2.24
+KoNViD_1k_videos/9724431856.mp4, 8.008008, 29.97, 3.84
+KoNViD_1k_videos/5460258278.mp4, 8.008, 29.97002997002997, 2.18
+KoNViD_1k_videos/7545358086.mp4, 8.0, 24.0, 2.02
+KoNViD_1k_videos/4206970561.mp4, 8.008, 23.976023976023978, 2.26
+KoNViD_1k_videos/10785115855.mp4, 8.008008, 29.97, 2.46
+KoNViD_1k_videos/9766188655.mp4, 8.008008, 29.97, 3.06
+KoNViD_1k_videos/12439507093.mp4, 8.0, 24.0, 2.5
+KoNViD_1k_videos/4582086726.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/6111574885.mp4, 8.008, 29.97002997002997, 2.74
+KoNViD_1k_videos/5346134105.mp4, 8.008, 29.97002997002997, 2.46
+KoNViD_1k_videos/3647017501.mp4, 8.008, 23.976023976023978, 3.56
+KoNViD_1k_videos/6660107279.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/8002170390.mp4, 8.0, 25.0, 2.76
+KoNViD_1k_videos/7495449440.mp4, 8.008, 29.97002997002997, 1.92
+KoNViD_1k_videos/6281135899.mp4, 8.008, 29.97002997002997, 3.0
+KoNViD_1k_videos/10124821225.mp4, 8.008008, 29.97, 3.4
+KoNViD_1k_videos/5004709068.mp4, 8.008, 23.976023976023978, 2.52
+KoNViD_1k_videos/8380926104.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/9494995600.mp4, 8.0, 24.0, 2.74
+KoNViD_1k_videos/5649370618.mp4, 8.008, 29.97002997002997, 2.9
+KoNViD_1k_videos/8110008288.mp4, 8.0, 24.0, 2.02
+KoNViD_1k_videos/4901210888.mp4, 8.008, 29.97002997002997, 2.82
+KoNViD_1k_videos/5356092090.mp4, 8.008, 29.97002997002997, 3.22
+KoNViD_1k_videos/4981160522.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/4236199952.mp4, 8.008, 29.97002997002997, 2.9
+KoNViD_1k_videos/7518773518.mp4, 8.008, 23.976023976023978, 4.32
+KoNViD_1k_videos/11902527296.mp4, 8.008, 29.97002997002997, 3.24
+KoNViD_1k_videos/6076608135.mp4, 8.008, 29.97002997002997, 4.4
+KoNViD_1k_videos/4020832533.mp4, 7.674333, 23.976023976023978, 2.948616601
+KoNViD_1k_videos/8587384006.mp4, 8.008, 29.97002997002997, 3.74
+KoNViD_1k_videos/8278672659.mp4, 8.008, 23.976023976023978, 2.18
+KoNViD_1k_videos/9586935224.mp4, 8.008008, 29.97, 3.34
+KoNViD_1k_videos/5357123491.mp4, 8.008, 23.976023976023978, 3.44
+KoNViD_1k_videos/5003283158.mp4, 8.008, 29.97002997002997, 2.78
+KoNViD_1k_videos/12682619933.mp4, 8.008008, 29.97, 3.24
+KoNViD_1k_videos/8217524401.mp4, 8.0, 24.0, 2.58
+KoNViD_1k_videos/10533906564.mp4, 8.008008, 29.97, 3.24
+KoNViD_1k_videos/4864754515.mp4, 8.0, 24.0, 3.12
+KoNViD_1k_videos/9267867823.mp4, 8.008008, 29.97, 3.84
+KoNViD_1k_videos/10636855406.mp4, 8.008008, 29.97, 3.88
+KoNViD_1k_videos/7819697028.mp4, 8.008, 29.97002997002997, 2.56
+KoNViD_1k_videos/6694553789.mp4, 8.008, 29.97002997002997, 3.62
+KoNViD_1k_videos/6982428911.mp4, 8.008, 29.97002997002997, 2.7
+KoNViD_1k_videos/6368379561.mp4, 8.008, 29.97002997002997, 3.54
+KoNViD_1k_videos/12623549414.mp4, 8.008008, 29.97, 3.68
+KoNViD_1k_videos/8442879599.mp4, 8.008, 29.97002997002997, 3.74
+KoNViD_1k_videos/8484812859.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/11816978255.mp4, 8.0, 25.0, 3.76
+KoNViD_1k_videos/8215074237.mp4, 8.0, 24.0, 2.7
+KoNViD_1k_videos/5415012877.mp4, 8.008, 23.976023976023978, 2.42
+KoNViD_1k_videos/12277706754.mp4, 8.008008, 29.97, 3.5
+KoNViD_1k_videos/11635304003.mp4, 8.008, 29.97002997002997, 1.74
+KoNViD_1k_videos/7526781566.mp4, 8.008, 29.97002997002997, 3.02
+KoNViD_1k_videos/6215542986.mp4, 8.008, 29.97002997002997, 2.58
+KoNViD_1k_videos/9363436148.mp4, 8.008008, 29.97, 3.58
+KoNViD_1k_videos/7005111650.mp4, 8.008, 23.976023976023978, 2.72
+KoNViD_1k_videos/5093919014.mp4, 8.008, 29.97002997002997, 2.26
+KoNViD_1k_videos/6671136435.mp4, 8.008, 29.97002997002997, 3.04
+KoNViD_1k_videos/3369925072.mp4, 8.008, 29.97002997002997, 4.02
+KoNViD_1k_videos/11057110163.mp4, 8.008, 29.97002997002997, 2.78
+KoNViD_1k_videos/3343535739.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/5526460224.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/5005236579.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/5203755303.mp4, 7.507499999999999, 29.97002997002997, 2.68
+KoNViD_1k_videos/5960200024.mp4, 8.008, 29.97002997002997, 2.52
+KoNViD_1k_videos/5453773681.mp4, 8.008, 29.97002997002997, 2.24
+KoNViD_1k_videos/6319757818.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/2999049224.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/5770178686.mp4, 8.008, 23.976023976023978, 3.78
+KoNViD_1k_videos/9737640625.mp4, 7.5492159999999995, 23.976, 3.36
+KoNViD_1k_videos/6107338487.mp4, 8.0, 25.0, 3.62
+KoNViD_1k_videos/5467704373.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/7897750494.mp4, 8.0, 24.0, 2.4
+KoNViD_1k_videos/4627984833.mp4, 8.008, 29.97002997002997, 1.7
+KoNViD_1k_videos/3611217365.mp4, 8.008, 23.976023976023978, 3.6
+KoNViD_1k_videos/5154464760.mp4, 8.0, 25.0, 3.86
+KoNViD_1k_videos/9802042845.mp4, 8.008008, 29.97, 3.66
+KoNViD_1k_videos/9618443862.mp4, 8.008008, 29.97, 2.08
+KoNViD_1k_videos/7892371292.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/4573314147.mp4, 8.008, 29.97002997002997, 2.56
+KoNViD_1k_videos/8026448058.mp4, 8.0, 25.0, 3.44
+KoNViD_1k_videos/4545845826.mp4, 8.008, 29.97002997002997, 3.9
+KoNViD_1k_videos/4567868868.mp4, 8.008, 29.97002997002997, 3.5
+KoNViD_1k_videos/6907839104.mp4, 8.008, 23.976023976023978, 3.5
+KoNViD_1k_videos/5875944717.mp4, 8.0, 25.0, 3.92
+KoNViD_1k_videos/5009379117.mp4, 7.8411669999999996, 29.97002997002997, 2.299595142
+KoNViD_1k_videos/8563986596.mp4, 8.008, 29.97002997002997, 2.94
+KoNViD_1k_videos/5987864177.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/4335304419.mp4, 8.008, 29.97002997002997, 3.02
+KoNViD_1k_videos/8098521993.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/11925275803.mp4, 8.008, 29.97002997002997, 2.66
+KoNViD_1k_videos/5956380805.mp4, 8.008, 29.97002997002997, 3.46
+KoNViD_1k_videos/4289215378.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/6787309364.mp4, 8.008, 29.97002997002997, 4.16
+KoNViD_1k_videos/10838934973.mp4, 8.008008, 29.97, 3.34
+KoNViD_1k_videos/7383994638.mp4, 8.008, 29.97002997002997, 3.74
+KoNViD_1k_videos/5845660280.mp4, 8.0, 25.0, 3.5
+KoNViD_1k_videos/8738388681.mp4, 8.008, 29.97002997002997, 3.8
+KoNViD_1k_videos/6210062524.mp4, 8.0, 24.0, 2.78
+KoNViD_1k_videos/6914710581.mp4, 8.0, 24.0, 4.0
+KoNViD_1k_videos/7151780093.mp4, 8.008, 29.97002997002997, 2.6
+KoNViD_1k_videos/11673027224.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/9317770917.mp4, 8.008008, 29.97, 3.38
+KoNViD_1k_videos/7548305462.mp4, 8.008, 23.976023976023978, 2.24
+KoNViD_1k_videos/4202404174.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/3425371223.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/4987165854.mp4, 8.008, 29.97002997002997, 3.24
+KoNViD_1k_videos/4642004278.mp4, 8.008, 29.97002997002997, 2.56
+KoNViD_1k_videos/6223104346.mp4, 8.008, 29.97002997002997, 2.56
+KoNViD_1k_videos/12088104156.mp4, 8.008, 29.97002997002997, 3.6
+KoNViD_1k_videos/5636101558.mp4, 8.0, 24.0, 2.26
+KoNViD_1k_videos/10238939436.mp4, 8.008008, 23.976, 2.58
+KoNViD_1k_videos/6844033217.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/8440795977.mp4, 8.008, 29.97002997002997, 2.28
+KoNViD_1k_videos/5570426341.mp4, 8.0, 24.0, 1.56
+KoNViD_1k_videos/9524637688.mp4, 8.008008, 29.97, 2.58
+KoNViD_1k_videos/3948863216.mp4, 8.0, 24.0, 3.62
+KoNViD_1k_videos/5706858082.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/5911601461.mp4, 8.008, 23.976023976023978, 3.42
+KoNViD_1k_videos/4836718842.mp4, 8.008, 29.97002997002997, 2.64
+KoNViD_1k_videos/6792979288.mp4, 8.0, 24.0, 1.78
+KoNViD_1k_videos/8486027818.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/12912266335.mp4, 8.0, 24.0, 2.86
+KoNViD_1k_videos/5893880541.mp4, 8.008, 23.976023976023978, 2.7
+KoNViD_1k_videos/8013211059.mp4, 8.0, 25.0, 2.48
+KoNViD_1k_videos/5440091902.mp4, 8.008, 29.97002997002997, 3.0
+KoNViD_1k_videos/5850199057.mp4, 8.008, 29.97002997002997, 2.4
+KoNViD_1k_videos/11082897705.mp4, 8.008008, 29.97, 3.3
+KoNViD_1k_videos/5338017301.mp4, 8.008, 29.97002997002997, 3.8
+KoNViD_1k_videos/7935377214.mp4, 8.008, 29.97002997002997, 3.58
+KoNViD_1k_videos/4871910775.mp4, 8.008, 29.97002997002997, 2.48
+KoNViD_1k_videos/8507613491.mp4, 8.008, 29.97002997002997, 2.28
+KoNViD_1k_videos/4321531777.mp4, 8.008, 29.97002997002997, 1.82
+KoNViD_1k_videos/8193379879.mp4, 8.0, 24.0, 3.26
+KoNViD_1k_videos/4712606231.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/6249802707.mp4, 8.008, 29.97002997002997, 3.58
+KoNViD_1k_videos/10283826203.mp4, 8.008008, 29.97, 4.14
+KoNViD_1k_videos/8435518163.mp4, 8.008, 29.97002997002997, 3.16
+KoNViD_1k_videos/8586743266.mp4, 8.008, 29.97002997002997, 2.72
+KoNViD_1k_videos/6490001399.mp4, 8.0, 24.0, 3.36
+KoNViD_1k_videos/11907021903.mp4, 8.008008, 29.97, 2.16
+KoNViD_1k_videos/7287459080.mp4, 8.008, 29.97002997002997, 2.32
+KoNViD_1k_videos/6065207696.mp4, 8.008, 29.97002997002997, 2.88
+KoNViD_1k_videos/7197025396.mp4, 8.008, 29.97002997002997, 3.92
+KoNViD_1k_videos/6092433561.mp4, 8.008, 29.97002997002997, 2.86
+KoNViD_1k_videos/5465225347.mp4, 7.5408669999999995, 29.97002997002997, 3.52
+KoNViD_1k_videos/3878700690.mp4, 8.008, 29.97002997002997, 2.74
+KoNViD_1k_videos/6313769652.mp4, 8.008, 29.97002997002997, 3.82
+KoNViD_1k_videos/5477809421.mp4, 8.008, 29.97002997002997, 4.04
+KoNViD_1k_videos/4639129836.mp4, 8.008, 29.97002997002997, 2.06
+KoNViD_1k_videos/7550224002.mp4, 8.0, 24.0, 2.34
+KoNViD_1k_videos/5805360253.mp4, 8.0, 24.0, 2.58
+KoNViD_1k_videos/6444142351.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/7177696763.mp4, 8.0, 24.0, 4.06
+KoNViD_1k_videos/4809572333.mp4, 8.008, 23.976023976023978, 3.12
+KoNViD_1k_videos/7664727482.mp4, 8.008, 29.97002997002997, 3.4
+KoNViD_1k_videos/7198853624.mp4, 8.008, 23.976023976023978, 3.6
+KoNViD_1k_videos/8124951690.mp4, 8.008, 29.97002997002997, 3.06
+KoNViD_1k_videos/12123666684.mp4, 8.008, 29.97002997002997, 4.0
+KoNViD_1k_videos/8132161360.mp4, 8.0, 24.0, 2.9
+KoNViD_1k_videos/7122092993.mp4, 8.008, 29.97002997002997, 2.88
+KoNViD_1k_videos/5625415555.mp4, 8.008, 23.976023976023978, 2.08
+KoNViD_1k_videos/6411380421.mp4, 8.008, 29.97002997002997, 3.44
+KoNViD_1k_videos/6867926575.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/8859509880.mp4, 8.008008, 29.97, 3.94
+KoNViD_1k_videos/5806274886.mp4, 8.0, 24.0, 2.24
+KoNViD_1k_videos/11425159465.mp4, 8.008008, 29.97, 2.68
+KoNViD_1k_videos/6648629457.mp4, 8.008, 23.976023976023978, 3.42
+KoNViD_1k_videos/8683385786.mp4, 8.008, 29.97002997002997, 3.66
+KoNViD_1k_videos/8701032421.mp4, 8.008, 23.976023976023978, 4.02
+KoNViD_1k_videos/7548202460.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/5030830413.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/5140332158.mp4, 8.008, 29.97002997002997, 3.58
+KoNViD_1k_videos/4723346787.mp4, 8.008, 29.97002997002997, 3.0
+KoNViD_1k_videos/5644303578.mp4, 8.008, 29.97002997002997, 2.54
+KoNViD_1k_videos/6807638360.mp4, 8.0, 24.0, 1.46
+KoNViD_1k_videos/5641862023.mp4, 8.008, 29.97002997002997, 3.36
+KoNViD_1k_videos/3860907449.mp4, 8.0, 24.0, 3.54
+KoNViD_1k_videos/5460445525.mp4, 8.0, 24.0, 2.04
+KoNViD_1k_videos/8339870239.mp4, 8.0, 24.0, 1.992063492
+KoNViD_1k_videos/7667993154.mp4, 8.008, 29.97002997002997, 3.76
+KoNViD_1k_videos/5869559051.mp4, 8.008, 29.97002997002997, 2.3
+KoNViD_1k_videos/13288073854.mp4, 8.0, 25.0, 4.1
+KoNViD_1k_videos/5360118581.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/5669052182.mp4, 8.008, 29.97002997002997, 3.879844961
+KoNViD_1k_videos/11971608074.mp4, 8.008, 29.97002997002997, 3.34
+KoNViD_1k_videos/8646170532.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/5786034294.mp4, 8.008, 23.976023976023978, 3.0
+KoNViD_1k_videos/7983610133.mp4, 8.008, 29.97002997002997, 3.74
+KoNViD_1k_videos/8084631711.mp4, 8.0, 24.0, 3.48
+KoNViD_1k_videos/11435365925.mp4, 8.008008, 29.97, 2.6
+KoNViD_1k_videos/6917745987.mp4, 8.008, 23.976023976023978, 2.4
+KoNViD_1k_videos/9209826160.mp4, 8.008008, 29.97, 3.34
+KoNViD_1k_videos/4907486229.mp4, 8.008, 29.97002997002997, 2.0
+KoNViD_1k_videos/6123040072.mp4, 8.0, 24.0, 2.22
+KoNViD_1k_videos/5023426388.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/5043852749.mp4, 7.8411669999999996, 29.97002997002997, 1.84
+KoNViD_1k_videos/10116604143.mp4, 8.008008, 23.976, 3.08
+KoNViD_1k_videos/4945025836.mp4, 8.0, 24.0, 3.32
+KoNViD_1k_videos/8664958871.mp4, 8.0, 25.0, 3.52
+KoNViD_1k_videos/5609569195.mp4, 8.008, 29.97002997002997, 2.82
+KoNViD_1k_videos/4931493432.mp4, 8.008, 29.97002997002997, 3.36
+KoNViD_1k_videos/9136867624.mp4, 8.008008, 29.97, 2.12
+KoNViD_1k_videos/8494495600.mp4, 8.008, 23.976023976023978, 3.74
+KoNViD_1k_videos/3781419322.mp4, 8.008, 23.976023976023978, 2.89787234
+KoNViD_1k_videos/8254380270.mp4, 8.0, 24.0, 3.16
+KoNViD_1k_videos/7863416600.mp4, 8.008, 29.97002997002997, 1.78
+KoNViD_1k_videos/4297792287.mp4, 8.0, 25.0, 3.78
+KoNViD_1k_videos/8510251919.mp4, 8.008, 29.97002997002997, 3.66
+KoNViD_1k_videos/4573734018.mp4, 8.008, 29.97002997002997, 3.3
+KoNViD_1k_videos/5562932712.mp4, 8.008, 23.976023976023978, 3.9
+KoNViD_1k_videos/7693986854.mp4, 8.008, 29.97002997002997, 3.1
+KoNViD_1k_videos/7959542546.mp4, 8.0, 24.0, 2.64
+KoNViD_1k_videos/7962972726.mp4, 8.008, 29.97002997002997, 2.18
+KoNViD_1k_videos/11618284403.mp4, 8.008, 29.97002997002997, 4.12
+KoNViD_1k_videos/6215522900.mp4, 8.008, 29.97002997002997, 3.48
+KoNViD_1k_videos/6804829908.mp4, 8.008, 29.97002997002997, 2.96
+KoNViD_1k_videos/6416687735.mp4, 8.008, 29.97002997002997, 3.36
+KoNViD_1k_videos/5003956763.mp4, 8.008, 29.97002997002997, 3.598455598
+KoNViD_1k_videos/3718963712.mp4, 8.0, 24.0, 3.66
+KoNViD_1k_videos/3482706595.mp4, 8.008, 29.97002997002997, 3.2
+KoNViD_1k_videos/6106291030.mp4, 8.008, 23.976023976023978, 3.6
+KoNViD_1k_videos/6847647728.mp4, 8.008, 29.97002997002997, 3.68
+KoNViD_1k_videos/7054450817.mp4, 8.0, 24.0, 3.46
+KoNViD_1k_videos/7558781186.mp4, 8.0, 25.0, 1.68
+KoNViD_1k_videos/12378508014.mp4, 8.0, 25.0, 4.035573123
+KoNViD_1k_videos/10132105415.mp4, 8.008008, 29.97, 1.64
+KoNViD_1k_videos/4975471705.mp4, 8.008, 29.97002997002997, 3.78
+KoNViD_1k_videos/5717841095.mp4, 8.008, 23.976023976023978, 3.02
+KoNViD_1k_videos/6212970517.mp4, 8.008, 29.97002997002997, 2.06
+KoNViD_1k_videos/9583034514.mp4, 8.008008, 29.97, 3.88
+KoNViD_1k_videos/10406095776.mp4, 8.008008, 29.97, 2.62
+KoNViD_1k_videos/8635322862.mp4, 8.008, 29.97002997002997, 2.36
+KoNViD_1k_videos/5354932311.mp4, 8.0, 24.0, 3.480314961
+KoNViD_1k_videos/5668545266.mp4, 8.0, 24.0, 2.46
+KoNViD_1k_videos/4201416279.mp4, 8.008, 29.97002997002997, 3.72
+KoNViD_1k_videos/8324243833.mp4, 8.008, 29.97002997002997, 2.16
+KoNViD_1k_videos/6439107119.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/8396734122.mp4, 8.0, 24.0, 2.84
+KoNViD_1k_videos/10027007645.mp4, 8.008008, 29.97, 3.8
+KoNViD_1k_videos/5176278265.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/3747645672.mp4, 8.0, 24.0, 3.82
+KoNViD_1k_videos/5127033299.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/6104508040.mp4, 8.008, 29.97002997002997, 3.08
+KoNViD_1k_videos/8632107216.mp4, 8.008, 29.97002997002997, 2.32
+KoNViD_1k_videos/4159445902.mp4, 8.008, 23.976023976023978, 2.52
+KoNViD_1k_videos/8933125503.mp4, 8.008008, 29.97, 2.4
+KoNViD_1k_videos/7198246948.mp4, 8.008, 29.97002997002997, 3.42
+KoNViD_1k_videos/4673672619.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/12143576166.mp4, 8.008008, 29.97, 3.44
+KoNViD_1k_videos/6079632868.mp4, 8.0, 25.0, 3.3
+KoNViD_1k_videos/8522230898.mp4, 8.008, 23.976023976023978, 3.34
+KoNViD_1k_videos/6916420878.mp4, 8.0, 24.0, 2.84
+KoNViD_1k_videos/8730797431.mp4, 8.008, 29.97002997002997, 3.7
+KoNViD_1k_videos/4735334918.mp4, 8.008, 29.97002997002997, 2.28
+KoNViD_1k_videos/6272702003.mp4, 8.008, 29.97002997002997, 3.5
+KoNViD_1k_videos/5311187352.mp4, 8.008, 29.97002997002997, 2.88
+KoNViD_1k_videos/13042523904.mp4, 8.008, 29.97002997002997, 3.12
+KoNViD_1k_videos/10691750555.mp4, 8.008008, 23.976, 2.84
+KoNViD_1k_videos/5539585646.mp4, 8.008, 29.97002997002997, 3.26
+KoNViD_1k_videos/6300963347.mp4, 8.008, 29.97002997002997, 3.58
+KoNViD_1k_videos/4329911079.mp4, 8.008, 29.97002997002997, 4.02
+KoNViD_1k_videos/7722155622.mp4, 8.0, 24.0, 2.36
+KoNViD_1k_videos/4491961485.mp4, 8.0, 24.0, 3.08
+KoNViD_1k_videos/7547764246.mp4, 8.008, 29.97002997002997, 3.28
+KoNViD_1k_videos/6041834062.mp4, 8.008, 29.97002997002997, 3.32
+KoNViD_1k_videos/6004183559.mp4, 8.008, 29.97002997002997, 3.38
+KoNViD_1k_videos/10128248563.mp4, 8.008008, 29.97, 2.86
+KoNViD_1k_videos/6012598017.mp4, 8.008, 29.97002997002997, 3.36
+KoNViD_1k_videos/6061346837.mp4, 8.008, 29.97002997002997, 2.78
+KoNViD_1k_videos/6416725133.mp4, 8.008, 29.97002997002997, 3.76
+KoNViD_1k_videos/8266961231.mp4, 8.008, 29.97002997002997, 3.14
+KoNViD_1k_videos/6868748154.mp4, 8.0, 25.0, 3.26
+KoNViD_1k_videos/5912268467.mp4, 8.008, 29.97002997002997, 3.34
+KoNViD_1k_videos/10404182556.mp4, 8.008008, 29.97, 2.32

examplar_data_labels/KoNiQ10k/test_labels.txt

@@ -0,0 +1,2015 @@
+10007357496.jpg, -1, -1, 68.7285714286
+10020766793.jpg, -1, -1, 81.50625
+10020891105.jpg, -1, -1, 56.8297413793
+10022757465.jpg, -1, -1, 71.0154867257
+10039534103.jpg, -1, -1, 76.075
+10050349236.jpg, -1, -1, 30.9418103448
+10086300493.jpg, -1, -1, 68.3026315789
+10093057603.jpg, -1, -1, 63.4955752212
+10113730664.jpg, -1, -1, 68.8125
+10126177964.jpg, -1, -1, 70.9078947368
+10154367094.jpg, -1, -1, 74.1120689655
+10154991453.jpg, -1, -1, 74.185483871
+10169304416.jpg, -1, -1, 80.8716814159
+10173544123.jpg, -1, -1, 65.1055555556
+10178183416.jpg, -1, -1, 75.6842105263
+10186832194.jpg, -1, -1, 57.5818584071
+10187102065.jpg, -1, -1, 38.7142857143
+10187522803.jpg, -1, -1, 68.4601769912
+10206674394.jpg, -1, -1, 79.1100917431
+10209917915.jpg, -1, -1, 46.9097222222
+10226524916.jpg, -1, -1, 20.6216216216
+10246318724.jpg, -1, -1, 74.3581081081
+10273804555.jpg, -1, -1, 68.3026315789
+10279904885.jpg, -1, -1, 44.0400943396
+1029470257.jpg, -1, -1, 43.0961538462
+10299438733.jpg, -1, -1, 55.2683486239
+10319186405.jpg, -1, -1, 55.5540540541
+10330831494.jpg, -1, -1, 48.5321100917
+10335072144.jpg, -1, -1, 62.1610576923
+10342569703.jpg, -1, -1, 38.6081081081
+10349956885.jpg, -1, -1, 69.75
+10350497833.jpg, -1, -1, 58.2605932203
+10351189716.jpg, -1, -1, 61.01875
+10362672006.jpg, -1, -1, 75.8189655172
+10366373284.jpg, -1, -1, 75.4545454545
+10397166423.jpg, -1, -1, 73.935840708
+10458703674.jpg, -1, -1, 38.1615044248
+10468110795.jpg, -1, -1, 61.4754464286
+10469774844.jpg, -1, -1, 59.6034482759
+10474002734.jpg, -1, -1, 80.7745535714
+10486107003.jpg, -1, -1, 67.9930555556
+10491423125.jpg, -1, -1, 72.5
+10520192995.jpg, -1, -1, 55.1630434783
+10526881085.jpg, -1, -1, 44.4598214286
+10536668436.jpg, -1, -1, 36.5432330827
+10539008983.jpg, -1, -1, 71.0154867257
+105620456.jpg, -1, -1, 55.8625
+10567114974.jpg, -1, -1, 64.847826087
+10582113466.jpg, -1, -1, 58.1607142857
+10603218265.jpg, -1, -1, 71.8616071429
+10625988224.jpg, -1, -1, 66.8443396226
+1063351510.jpg, -1, -1, 47.1153846154
+10633933603.jpg, -1, -1, 58.8898305085
+10644234544.jpg, -1, -1, 54.7289915966
+106652866.jpg, -1, -1, 49.0441176471
+1067765170.jpg, -1, -1, 51.175
+10678350303.jpg, -1, -1, 66.786637931
+10680715605.jpg, -1, -1, 37.3
+10680932454.jpg, -1, -1, 51.39375
+10689376076.jpg, -1, -1, 59.2394067797
+10714662986.jpg, -1, -1, 80.6551724138
+10715577856.jpg, -1, -1, 45.213592233
+10721124556.jpg, -1, -1, 58.4579646018
+10740727595.jpg, -1, -1, 72.9125
+10762659033.jpg, -1, -1, 42.25
+10767441335.jpg, -1, -1, 74.2094594595
+10770700455.jpg, -1, -1, 60.74375
+10773126784.jpg, -1, -1, 78.475
+10776111906.jpg, -1, -1, 73.975
+10791502435.jpg, -1, -1, 70.8130252101
+10813730303.jpg, -1, -1, 40.0817307692
+10836905604.jpg, -1, -1, 75.9932432432
+1084984228.jpg, -1, -1, 34.8319327731
+10861031374.jpg, -1, -1, 64.675
+10865798996.jpg, -1, -1, 76.4722222222
+10866463396.jpg, -1, -1, 69.1029411765
+10880314705.jpg, -1, -1, 77.4210526316
+10911046243.jpg, -1, -1, 76.2339449541
+10912602716.jpg, -1, -1, 74.9510869565
+10953973716.jpg, -1, -1, 72.6486486486
+11013892854.jpg, -1, -1, 27.4978813559
+110356262.jpg, -1, -1, 35.925
+1103755154.jpg, -1, -1, 73.0641025641
+11051374673.jpg, -1, -1, 62.6148648649
+11054638863.jpg, -1, -1, 61.0070754717
+11054899145.jpg, -1, -1, 74.0512820513
+11060185873.jpg, -1, -1, 63.0148305085
+11095381506.jpg, -1, -1, 76.75
+11109548195.jpg, -1, -1, 61.5231092437
+11126652126.jpg, -1, -1, 56.6485849057
+11150131936.jpg, -1, -1, 63.9158878505
+11153943035.jpg, -1, -1, 60.2777777778
+11156365526.jpg, -1, -1, 47.5882352941
+11158314215.jpg, -1, -1, 46.3349514563
+11215941184.jpg, -1, -1, 73.4077669903
+11232665444.jpg, -1, -1, 32.2668067227
+11242015725.jpg, -1, -1, 63.7152777778
+11250215644.jpg, -1, -1, 69.6776315789
+11253432804.jpg, -1, -1, 61.2027027027
+11260270773.jpg, -1, -1, 72.3513513514
+11293176185.jpg, -1, -1, 61.3
+11310848503.jpg, -1, -1, 70.1461864407
+11335231214.jpg, -1, -1, 68.9603960396
+113663782.jpg, -1, -1, 67.55
+11378848184.jpg, -1, -1, 64.1334745763
+11406584295.jpg, -1, -1, 55.3310810811
+11407368256.jpg, -1, -1, 72.2884615385
+11407862453.jpg, -1, -1, 70.6504424779
+11408443745.jpg, -1, -1, 63.475
+11433127464.jpg, -1, -1, 54.109375
+11436390143.jpg, -1, -1, 72.4481132075
+11436508006.jpg, -1, -1, 65.14375
+11438612393.jpg, -1, -1, 31.9953271028
+11446005074.jpg, -1, -1, 77.468487395
+11452299875.jpg, -1, -1, 74.4608695652
+11476172786.jpg, -1, -1, 72.625
+11511265293.jpg, -1, -1, 24.2767857143
+11514342335.jpg, -1, -1, 71.9646017699
+11522920033.jpg, -1, -1, 57.3275862069
+11527524903.jpg, -1, -1, 33.6972477064
+115290990.jpg, -1, -1, 42.9432773109
+1159028861.jpg, -1, -1, 72.9125
+11706252.jpg, -1, -1, 32.2521186441
+121123359.jpg, -1, -1, 88.24375
+122760647.jpg, -1, -1, 50.9125
+123242995.jpg, -1, -1, 60.968487395
+128079586.jpg, -1, -1, 57.6407563025
+1289162225.jpg, -1, -1, 62.1470588235
+1313706108.jpg, -1, -1, 34.4159663866
+1329713028.jpg, -1, -1, 52.7632743363
+1331642031.jpg, -1, -1, 54.2905405405
+1348094685.jpg, -1, -1, 72.3834745763
+139999071.jpg, -1, -1, 25.68125
+1408118135.jpg, -1, -1, 44.5132743363
+143394138.jpg, -1, -1, 29.94375
+143622689.jpg, -1, -1, 62.7375
+1465282532.jpg, -1, -1, 35.6941964286
+1465706268.jpg, -1, -1, 59.78125
+1519650440.jpg, -1, -1, 53.156779661
+154270569.jpg, -1, -1, 70.4772727273
+156266461.jpg, -1, -1, 41.0125
+1571972975.jpg, -1, -1, 50.9004854369
+159144665.jpg, -1, -1, 74.15
+16079437.jpg, -1, -1, 53.0868644068
+163208499.jpg, -1, -1, 27.3181818182
+166528976.jpg, -1, -1, 62.25625
+16799807.jpg, -1, -1, 39.9030172414
+1748861184.jpg, -1, -1, 62.2457627119
+175326787.jpg, -1, -1, 35.2272727273
+17606027.jpg, -1, -1, 69.3863636364
+1794888173.jpg, -1, -1, 49.5911016949
+1797683075.jpg, -1, -1, 34.8897058824
+1802306299.jpg, -1, -1, 40.2236842105
+1807195948.jpg, -1, -1, 34.0
+1809848435.jpg, -1, -1, 24.7278761062
+1812202754.jpg, -1, -1, 78.4661016949
+18438855.jpg, -1, -1, 74.6207627119
+187499391.jpg, -1, -1, 46.8256302521
+1946326263.jpg, -1, -1, 67.0736607143
+1970440733.jpg, -1, -1, 55.875
+1988371580.jpg, -1, -1, 68.4935344828
+1990074154.jpg, -1, -1, 82.4868421053
+19943010.jpg, -1, -1, 77.2605042017
+2035975268.jpg, -1, -1, 15.1428571429
+203950866.jpg, -1, -1, 72.7542016807
+2041755502.jpg, -1, -1, 74.1064356436
+2055589084.jpg, -1, -1, 38.4806034483
+2059981058.jpg, -1, -1, 58.75
+2065722620.jpg, -1, -1, 67.9662162162
+2070418201.jpg, -1, -1, 46.8109756098
+2074092022.jpg, -1, -1, 68.2866972477
+2088433684.jpg, -1, -1, 33.8526785714
+2102208013.jpg, -1, -1, 81.017699115
+211603206.jpg, -1, -1, 65.474789916
+2126677316.jpg, -1, -1, 29.3990384615
+2138874780.jpg, -1, -1, 70.4779411765
+213906996.jpg, -1, -1, 86.6625
+2140758242.jpg, -1, -1, 69.0817757009
+2150291485.jpg, -1, -1, 50.8650442478
+2161809899.jpg, -1, -1, 57.7924107143
+2168323029.jpg, -1, -1, 53.05
+2181528959.jpg, -1, -1, 71.1949152542
+2191549592.jpg, -1, -1, 50.7682926829
+2191619969.jpg, -1, -1, 59.575
+2195703133.jpg, -1, -1, 52.3982300885
+2199234449.jpg, -1, -1, 45.2894736842
+2211658928.jpg, -1, -1, 72.7873134328
+2215695798.jpg, -1, -1, 63.0172413793
+2223002938.jpg, -1, -1, 63.2569444444
+2224984425.jpg, -1, -1, 67.4459459459
+2234064642.jpg, -1, -1, 44.125
+2235352888.jpg, -1, -1, 70.5774336283
+2236324282.jpg, -1, -1, 72.5
+2239987680.jpg, -1, -1, 29.9121621622
+224076897.jpg, -1, -1, 62.2857142857
+2243706700.jpg, -1, -1, 53.3986486486
+2255948084.jpg, -1, -1, 74.8644859813
+2266672191.jpg, -1, -1, 57.30625
+2280116666.jpg, -1, -1, 34.0711206897
+2280668053.jpg, -1, -1, 48.0907079646
+2285789022.jpg, -1, -1, 51.6
+2298963655.jpg, -1, -1, 72.5514018692
+2301461383.jpg, -1, -1, 57.25
+2313971368.jpg, -1, -1, 69.891025641
+2322493147.jpg, -1, -1, 68.0951327434
+2337030121.jpg, -1, -1, 74.2837837838
+2338832331.jpg, -1, -1, 55.5338983051
+234944177.jpg, -1, -1, 72.56875
+2349676135.jpg, -1, -1, 37.025
+2350031609.jpg, -1, -1, 40.0
+2362202039.jpg, -1, -1, 75.6730769231
+2366273919.jpg, -1, -1, 29.5066964286
+2372696116.jpg, -1, -1, 62.0353982301
+2383953966.jpg, -1, -1, 65.5526315789
+2409675838.jpg, -1, -1, 37.1890756303
+2415334988.jpg, -1, -1, 38.4094827586
+2418067534.jpg, -1, -1, 32.1281512605
+2422029745.jpg, -1, -1, 27.0486111111
+2448776548.jpg, -1, -1, 56.75625
+2459843879.jpg, -1, -1, 63.7
+2466673147.jpg, -1, -1, 69.5603448276
+2474265369.jpg, -1, -1, 40.6717391304
+2486040810.jpg, -1, -1, 57.5304347826
+2491353954.jpg, -1, -1, 47.2876106195
+2492228578.jpg, -1, -1, 31.1756756757
+2510458503.jpg, -1, -1, 74.3173913043
+2517718928.jpg, -1, -1, 73.0681818182
+2539818768.jpg, -1, -1, 40.1707317073
+2549004146.jpg, -1, -1, 41.7740384615
+2555644998.jpg, -1, -1, 66.55
+2557729909.jpg, -1, -1, 77.1280487805
+2568268182.jpg, -1, -1, 47.5586956522
+257457837.jpg, -1, -1, 38.74375
+2579041082.jpg, -1, -1, 63.5096153846
+2585752058.jpg, -1, -1, 58.6728971963
+2604683646.jpg, -1, -1, 58.371559633
+2612353606.jpg, -1, -1, 79.375
+2620467173.jpg, -1, -1, 53.6025641026
+2622466873.jpg, -1, -1, 57.9174311927
+2624809780.jpg, -1, -1, 57.2375
+2624882684.jpg, -1, -1, 70.0504201681
+2627264031.jpg, -1, -1, 74.5810810811
+2640025600.jpg, -1, -1, 47.2119565217
+2652397944.jpg, -1, -1, 70.3
+2657101675.jpg, -1, -1, 61.5504587156
+2659313316.jpg, -1, -1, 71.9345794393
+2676837522.jpg, -1, -1, 60.7720588235
+2679224314.jpg, -1, -1, 17.01875
+268066639.jpg, -1, -1, 54.1666666667
+2700852747.jpg, -1, -1, 55.511682243
+2713735764.jpg, -1, -1, 51.5683453237
+2720757001.jpg, -1, -1, 32.8541666667
+2722913532.jpg, -1, -1, 60.7935779817
+2725531194.jpg, -1, -1, 68.1785714286
+273181633.jpg, -1, -1, 68.6792035398
+2737304398.jpg, -1, -1, 53.875
+273844105.jpg, -1, -1, 79.7368421053
+2739007601.jpg, -1, -1, 51.2071428571
+274390572.jpg, -1, -1, 36.5169491525
+2746983219.jpg, -1, -1, 67.5789473684
+2748370097.jpg, -1, -1, 56.0252293578
+276590560.jpg, -1, -1, 40.1689189189
+2766412345.jpg, -1, -1, 37.1216216216
+2778039171.jpg, -1, -1, 68.8581081081
+2779185055.jpg, -1, -1, 49.9889380531
+2796774323.jpg, -1, -1, 78.4174107143
+2798808586.jpg, -1, -1, 62.1559633028
+2801033102.jpg, -1, -1, 43.4978991597
+2807762234.jpg, -1, -1, 70.4663865546
+2808286292.jpg, -1, -1, 72.3552631579
+28173586.jpg, -1, -1, 63.5862068966
+2817883212.jpg, -1, -1, 79.7695652174
+2818844013.jpg, -1, -1, 71.0471698113
+2819296433.jpg, -1, -1, 46.1956521739
+2839494557.jpg, -1, -1, 71.4369747899
+2844802224.jpg, -1, -1, 52.5067567568
+2851510265.jpg, -1, -1, 63.1399082569
+2852195444.jpg, -1, -1, 64.3509174312
+2856053454.jpg, -1, -1, 26.95
+2861346409.jpg, -1, -1, 62.5344036697
+2866097268.jpg, -1, -1, 46.0193965517
+2867718050.jpg, -1, -1, 31.86875
+2876226509.jpg, -1, -1, 38.3676470588
+2889734029.jpg, -1, -1, 74.6260504202
+289462174.jpg, -1, -1, 26.9695652174
+290161904.jpg, -1, -1, 68.925
+2907072293.jpg, -1, -1, 21.1666666667
+2910782740.jpg, -1, -1, 56.1155462185
+2932540851.jpg, -1, -1, 67.6239495798
+2943372707.jpg, -1, -1, 64.2033898305
+2952908030.jpg, -1, -1, 50.2179487179
+2954406623.jpg, -1, -1, 26.3739495798
+2960552821.jpg, -1, -1, 23.9565217391
+2962584804.jpg, -1, -1, 61.3326086957
+2964746234.jpg, -1, -1, 38.725
+2967177700.jpg, -1, -1, 52.4888392857
+296739390.jpg, -1, -1, 55.0833333333
+2974085281.jpg, -1, -1, 42.9551282051
+2976379229.jpg, -1, -1, 75.9491525424
+2977820017.jpg, -1, -1, 62.5855263158
+2982904736.jpg, -1, -1, 74.4469026549
+298392087.jpg, -1, -1, 64.2242990654
+2985434942.jpg, -1, -1, 49.125
+301104017.jpg, -1, -1, 72.43125
+3011247965.jpg, -1, -1, 55.2567567568
+3013640771.jpg, -1, -1, 70.9776785714
+3018181041.jpg, -1, -1, 49.2086956522
+3021659624.jpg, -1, -1, 53.18125
+3022463774.jpg, -1, -1, 71.1596638655
+3026154930.jpg, -1, -1, 56.61875
+302681636.jpg, -1, -1, 53.5762711864
+3050837263.jpg, -1, -1, 53.7702702703
+3070894795.jpg, -1, -1, 37.743697479
+3072804517.jpg, -1, -1, 63.1176470588
+3075864902.jpg, -1, -1, 19.0945378151
+3077228870.jpg, -1, -1, 34.3055555556
+3084516747.jpg, -1, -1, 30.6153846154
+3084860790.jpg, -1, -1, 12.1850961538
+3087855667.jpg, -1, -1, 45.5202702703
+3090534929.jpg, -1, -1, 37.0320512821
+3093435000.jpg, -1, -1, 73.4474789916
+310467976.jpg, -1, -1, 75.7352941176
+3117729596.jpg, -1, -1, 50.6513761468
+3118439925.jpg, -1, -1, 62.2432432432
+3118885310.jpg, -1, -1, 62.4428571429
+311996877.jpg, -1, -1, 45.0924369748
+3130211082.jpg, -1, -1, 76.7697368421
+3132085493.jpg, -1, -1, 68.4807692308
+3132669524.jpg, -1, -1, 28.5924369748
+3136368482.jpg, -1, -1, 46.6399082569
+3141055360.jpg, -1, -1, 54.6973684211
+3147927973.jpg, -1, -1, 55.9075630252
+3148698494.jpg, -1, -1, 47.1697247706
+3164597811.jpg, -1, -1, 68.3357142857
+3164921799.jpg, -1, -1, 57.2767857143
+3174536924.jpg, -1, -1, 54.7648305085
+3175529254.jpg, -1, -1, 55.873853211
+3177560540.jpg, -1, -1, 61.3
+3177839887.jpg, -1, -1, 78.4410377358
+3184407773.jpg, -1, -1, 39.6816037736
+3184831791.jpg, -1, -1, 71.1621621622
+3188292993.jpg, -1, -1, 77.9256756757
+3194036363.jpg, -1, -1, 67.15
+3197293330.jpg, -1, -1, 65.8786407767
+3206894685.jpg, -1, -1, 56.9428571429
+3207112247.jpg, -1, -1, 58.1659292035
+320987228.jpg, -1, -1, 83.4306722689
+3210385990.jpg, -1, -1, 70.3846153846
+3217227849.jpg, -1, -1, 48.6666666667
+3224295308.jpg, -1, -1, 61.0334821429
+3233213622.jpg, -1, -1, 65.474789916
+3241749396.jpg, -1, -1, 58.5943396226
+3243501762.jpg, -1, -1, 37.3445945946
+3252078614.jpg, -1, -1, 47.81875
+325223545.jpg, -1, -1, 67.1434782609
+3253636665.jpg, -1, -1, 77.79375
+3254432034.jpg, -1, -1, 60.5525210084
+3256723994.jpg, -1, -1, 54.9933035714
+326195053.jpg, -1, -1, 50.9194915254
+3265568572.jpg, -1, -1, 70.1025641026
+3275523745.jpg, -1, -1, 57.5304347826
+3277186380.jpg, -1, -1, 54.2754237288
+3296835726.jpg, -1, -1, 66.0337837838
+3297887749.jpg, -1, -1, 69.608974359
+3301587792.jpg, -1, -1, 43.4174528302
+3302975736.jpg, -1, -1, 57.7456521739
+3303080079.jpg, -1, -1, 80.527027027
+3305968848.jpg, -1, -1, 23.7064220183
+3307215666.jpg, -1, -1, 50.5705128205
+3308341801.jpg, -1, -1, 27.8474576271
+3311728409.jpg, -1, -1, 75.4615384615
+3312331811.jpg, -1, -1, 18.9601769912
+3315383233.jpg, -1, -1, 21.2905405405
+3323033903.jpg, -1, -1, 26.2702702703
+3326988489.jpg, -1, -1, 57.375
+3330470420.jpg, -1, -1, 54.6628440367
+3335618194.jpg, -1, -1, 75.4101941748
+3338894282.jpg, -1, -1, 63.7311320755
+3340734956.jpg, -1, -1, 63.7995689655
+3341085556.jpg, -1, -1, 20.6032110092
+3346416285.jpg, -1, -1, 57.2362385321
+3347823880.jpg, -1, -1, 67.0
+3355200103.jpg, -1, -1, 75.3645833333
+3355464174.jpg, -1, -1, 60.3282608696
+3362731032.jpg, -1, -1, 76.69375
+3364722083.jpg, -1, -1, 44.5334821429
+3376076568.jpg, -1, -1, 56.9625
+3381345467.jpg, -1, -1, 46.6305309735
+3383701765.jpg, -1, -1, 65.6621621622
+3388712525.jpg, -1, -1, 48.1216216216
+3390605400.jpg, -1, -1, 25.0688073394
+342221498.jpg, -1, -1, 32.3361344538
+3427902428.jpg, -1, -1, 71.2983193277
+3427902514.jpg, -1, -1, 69.3130841121
+3433315492.jpg, -1, -1, 49.6238938053
+3444483519.jpg, -1, -1, 66.1680672269
+3446366493.jpg, -1, -1, 67.2313084112
+3450753884.jpg, -1, -1, 69.2632743363
+3452963187.jpg, -1, -1, 25.3040540541
+3456840631.jpg, -1, -1, 76.8214285714
+3458045907.jpg, -1, -1, 60.5178571429
+3461062752.jpg, -1, -1, 44.5540540541
+3462395913.jpg, -1, -1, 49.5424107143
+3463146453.jpg, -1, -1, 68.6805555556
+3463376543.jpg, -1, -1, 56.3055555556
+3466610103.jpg, -1, -1, 70.3559322034
+3469092597.jpg, -1, -1, 68.1148648649
+347162205.jpg, -1, -1, 31.5425531915
+34796696.jpg, -1, -1, 66.1538461538
+348162485.jpg, -1, -1, 26.3046218487
+3483143101.jpg, -1, -1, 6.06092436975
+3483540691.jpg, -1, -1, 46.1891891892
+3487085790.jpg, -1, -1, 69.0442477876
+3487701269.jpg, -1, -1, 69.5631067961
+3489731995.jpg, -1, -1, 66.0043103448
+3497202348.jpg, -1, -1, 55.5540540541
+35002160.jpg, -1, -1, 68.468220339
+3503300231.jpg, -1, -1, 76.41875
+3505262206.jpg, -1, -1, 50.4269911504
+3505627590.jpg, -1, -1, 40.703125
+3510063780.jpg, -1, -1, 46.052173913
+3515663938.jpg, -1, -1, 73.4821428571
+351997474.jpg, -1, -1, 50.6586538462
+352602458.jpg, -1, -1, 29.552173913
+352744670.jpg, -1, -1, 14.3783783784
+353228664.jpg, -1, -1, 66.0508849558
+3532374814.jpg, -1, -1, 68.8245192308
+3533949779.jpg, -1, -1, 63.6415929204
+3537044665.jpg, -1, -1, 34.2796610169
+3542002438.jpg, -1, -1, 47.0336134454
+3547073718.jpg, -1, -1, 77.2942477876
+3563134047.jpg, -1, -1, 70.0353773585
+357641600.jpg, -1, -1, 21.3130252101
+358433606.jpg, -1, -1, 27.0607476636
+3585032307.jpg, -1, -1, 48.4571428571
+3587076357.jpg, -1, -1, 62.9108695652
+3587337138.jpg, -1, -1, 44.5862831858
+3596611917.jpg, -1, -1, 54.4152542373
+3597095413.jpg, -1, -1, 76.0588235294
+3597199001.jpg, -1, -1, 71.3347457627
+359723983.jpg, -1, -1, 72.0934782609
+3606669688.jpg, -1, -1, 60.2831858407
+3608137140.jpg, -1, -1, 70.4932432432
+3609473273.jpg, -1, -1, 77.4358974359
+3615125120.jpg, -1, -1, 67.6991525424
+3615267731.jpg, -1, -1, 70.3584070796
+3616806128.jpg, -1, -1, 63.3669724771
+3618474326.jpg, -1, -1, 69.0810810811
+3618795157.jpg, -1, -1, 83.3443396226
+3620726.jpg, -1, -1, 24.224789916
+3632417985.jpg, -1, -1, 4.16741071429
+3639490630.jpg, -1, -1, 46.2375
+3645793167.jpg, -1, -1, 71.7257281553
+3648316650.jpg, -1, -1, 65.0321100917
+3651951998.jpg, -1, -1, 56.8276699029
+3663320781.jpg, -1, -1, 75.9612068966
+3663960871.jpg, -1, -1, 56.856557377
+3665983083.jpg, -1, -1, 78.1486486486
+3671193503.jpg, -1, -1, 60.1621621622
+3674117315.jpg, -1, -1, 61.5723684211
+3680743288.jpg, -1, -1, 34.6292372881
+3683418770.jpg, -1, -1, 55.9533898305
+3689126503.jpg, -1, -1, 74.4174311927
+3705698114.jpg, -1, -1, 53.7123893805
+3705884386.jpg, -1, -1, 46.2673913043
+3712529193.jpg, -1, -1, 36.7037815126
+3713010621.jpg, -1, -1, 44.0178571429
+3713773458.jpg, -1, -1, 64.7572815534
+3715014568.jpg, -1, -1, 48.9391891892
+3721179887.jpg, -1, -1, 43.5878378378
+3722409963.jpg, -1, -1, 34.1386554622
+3722623767.jpg, -1, -1, 51.1991525424
+3727572481.jpg, -1, -1, 52.5817757009
+3733759776.jpg, -1, -1, 68.9888392857
+3734577194.jpg, -1, -1, 66.6108490566
+3736358849.jpg, -1, -1, 31.8
+3736907392.jpg, -1, -1, 65.3207964602
+375084308.jpg, -1, -1, 59.0527522936
+3753143338.jpg, -1, -1, 11.51875
+3755358004.jpg, -1, -1, 69.3040540541
+3758843972.jpg, -1, -1, 64.343220339
+3759763464.jpg, -1, -1, 48.5776699029
+3761173384.jpg, -1, -1, 53.4464285714
+3767341492.jpg, -1, -1, 35.275
+3767933394.jpg, -1, -1, 75.0943396226
+3783245891.jpg, -1, -1, 52.5974576271
+3783568353.jpg, -1, -1, 60.6152173913
+3788379941.jpg, -1, -1, 74.8849557522
+379437662.jpg, -1, -1, 68.6651376147
+38056572.jpg, -1, -1, 64.8326271186
+3810011859.jpg, -1, -1, 36.497706422
+3812031514.jpg, -1, -1, 70.9851694915
+3812926720.jpg, -1, -1, 70.3302752294
+3815063597.jpg, -1, -1, 47.3325892857
+3815098207.jpg, -1, -1, 74.7297297297
+3817779484.jpg, -1, -1, 50.7426470588
+3824413541.jpg, -1, -1, 76.3879310345
+3825837186.jpg, -1, -1, 70.0320512821
+3827560087.jpg, -1, -1, 73.0260869565
+3839953318.jpg, -1, -1, 57.281779661
+3841233055.jpg, -1, -1, 43.0981308411
+3841305755.jpg, -1, -1, 61.3368644068
+3845656628.jpg, -1, -1, 62.5855263158
+3847965007.jpg, -1, -1, 43.0
+3851672801.jpg, -1, -1, 75.9189189189
+3851941327.jpg, -1, -1, 55.495412844
+3853359984.jpg, -1, -1, 61.4471153846
+3854279335.jpg, -1, -1, 30.5558252427
+3855419676.jpg, -1, -1, 76.2905405405
+3860848792.jpg, -1, -1, 43.6585365854
+3867441709.jpg, -1, -1, 65.9048672566
+386746008.jpg, -1, -1, 62.4243697479
+387003357.jpg, -1, -1, 31.7702702703
+3870343367.jpg, -1, -1, 62.4705882353
+3871801313.jpg, -1, -1, 50.6398305085
+3874298550.jpg, -1, -1, 61.84375
+3874382715.jpg, -1, -1, 32.5721153846
+3880140639.jpg, -1, -1, 34.5840707965
+3890811565.jpg, -1, -1, 64.1575630252
+3895433023.jpg, -1, -1, 58.1760869565
+390226070.jpg, -1, -1, 72.2566371681
+3905518312.jpg, -1, -1, 39.3035714286
+3908866899.jpg, -1, -1, 38.2777777778
+3917849504.jpg, -1, -1, 49.4956521739
+3919380535.jpg, -1, -1, 69.1361607143
+392103652.jpg, -1, -1, 33.3707627119
+3921589638.jpg, -1, -1, 72.8407079646
+3922421993.jpg, -1, -1, 72.0416666667
+3930941584.jpg, -1, -1, 36.4527027027
+3931210821.jpg, -1, -1, 55.2567567568
+3934821151.jpg, -1, -1, 69.5683962264
+3935177366.jpg, -1, -1, 21.0256410256
+3938367619.jpg, -1, -1, 50.2155172414
+3949143151.jpg, -1, -1, 68.99375
+395315549.jpg, -1, -1, 52.1875
+39540491.jpg, -1, -1, 73.3088235294
+3960444662.jpg, -1, -1, 73.7375
+396221760.jpg, -1, -1, 61.5948275862
+3969622447.jpg, -1, -1, 45.3737864078
+3975469304.jpg, -1, -1, 82.3851351351
+3980206588.jpg, -1, -1, 63.6116071429
+3983499964.jpg, -1, -1, 67.1542056075
+3984248064.jpg, -1, -1, 24.602173913
+3985027264.jpg, -1, -1, 40.0260869565
+3985258920.jpg, -1, -1, 52.3877118644
+3993808694.jpg, -1, -1, 40.875
+3999106249.jpg, -1, -1, 50.2828947368
+4003392015.jpg, -1, -1, 37.7905405405
+4004417944.jpg, -1, -1, 30.8475609756
+4016051317.jpg, -1, -1, 32.3348623853
+4018386659.jpg, -1, -1, 58.0688073394
+4018879156.jpg, -1, -1, 38.2867647059
+4028266820.jpg, -1, -1, 65.9423076923
+4039710407.jpg, -1, -1, 70.85
+4040650709.jpg, -1, -1, 45.0
+4042188530.jpg, -1, -1, 39.4495412844
+4042433513.jpg, -1, -1, 49.4976635514
+4042572733.jpg, -1, -1, 41.4324324324
+4054847919.jpg, -1, -1, 76.9951923077
+4079545582.jpg, -1, -1, 58.8243243243
+4081025902.jpg, -1, -1, 43.9868421053
+4089514101.jpg, -1, -1, 74.1548672566
+4094850958.jpg, -1, -1, 29.2638888889
+4095160762.jpg, -1, -1, 65.8318584071
+409597917.jpg, -1, -1, 40.251953125
+4098700616.jpg, -1, -1, 65.5897435897
+4100096331.jpg, -1, -1, 18.5405405405
+4107362083.jpg, -1, -1, 80.75
+4109408743.jpg, -1, -1, 73.9915254237
+411025625.jpg, -1, -1, 71.8931034483
+411090819.jpg, -1, -1, 16.9
+4113685341.jpg, -1, -1, 7.51680672269
+4126410943.jpg, -1, -1, 40.0065789474
+4132747677.jpg, -1, -1, 62.6458333333
+4139461254.jpg, -1, -1, 66.2142857143
+4139991387.jpg, -1, -1, 56.7413043478
+414385138.jpg, -1, -1, 68.7408256881
+4144540155.jpg, -1, -1, 62.9121621622
+4147791219.jpg, -1, -1, 58.2102803738
+4149265699.jpg, -1, -1, 66.6071428571
+4156345687.jpg, -1, -1, 75.193877551
+4156652933.jpg, -1, -1, 62.5576923077
+4157689684.jpg, -1, -1, 47.1415929204
+4167927417.jpg, -1, -1, 48.3846153846
+4169643886.jpg, -1, -1, 48.9111111111
+4172609703.jpg, -1, -1, 39.0769230769
+4172859607.jpg, -1, -1, 57.4121621622
+4172951303.jpg, -1, -1, 56.1216814159
+4176378029.jpg, -1, -1, 47.95
+4178432884.jpg, -1, -1, 48.4551282051
+4185039279.jpg, -1, -1, 72.7542016807
+4187118157.jpg, -1, -1, 76.1034482759
+4191882783.jpg, -1, -1, 45.025
+4196316000.jpg, -1, -1, 31.73125
+4206098907.jpg, -1, -1, 43.7867647059
+4213244648.jpg, -1, -1, 66.9282608696
+4224633098.jpg, -1, -1, 80.2146017699
+4236732050.jpg, -1, -1, 67.5606796117
+4239526963.jpg, -1, -1, 72.43125
+4242548000.jpg, -1, -1, 64.7572815534
+4243562466.jpg, -1, -1, 67.1398305085
+4243875190.jpg, -1, -1, 69.1949541284
+4244935320.jpg, -1, -1, 64.1575630252
+42490373.jpg, -1, -1, 62.9461206897
+4255841453.jpg, -1, -1, 61.0540540541
+4256128407.jpg, -1, -1, 31.4891304348
+4257293580.jpg, -1, -1, 26.9978991597
+4259043215.jpg, -1, -1, 64.5529661017
+4260146768.jpg, -1, -1, 69.6490825688
+4264601430.jpg, -1, -1, 73.8869565217
+4265877256.jpg, -1, -1, 50.2268211921
+4275452644.jpg, -1, -1, 66.8526785714
+4281583432.jpg, -1, -1, 62.66875
+4286953087.jpg, -1, -1, 35.3983050847
+4296647005.jpg, -1, -1, 72.3428571429
+4297599043.jpg, -1, -1, 63.7
+4298067776.jpg, -1, -1, 72.7
+4310499723.jpg, -1, -1, 52.5086956522
+4312794040.jpg, -1, -1, 62.9789915966
+4313162470.jpg, -1, -1, 67.2079831933
+4315078660.jpg, -1, -1, 41.6207627119
+4316201498.jpg, -1, -1, 67.2946428571
+4323755897.jpg, -1, -1, 61.4256756757
+4324984866.jpg, -1, -1, 63.3419811321
+433731087.jpg, -1, -1, 24.325
+434127863.jpg, -1, -1, 58.3080357143
+4346999174.jpg, -1, -1, 72.373853211
+4353430545.jpg, -1, -1, 27.2697368421
+4353485101.jpg, -1, -1, 32.3938053097
+4359682859.jpg, -1, -1, 78.1913043478
+4360105019.jpg, -1, -1, 29.4652317881
+43749901.jpg, -1, -1, 37.3559322034
+4378419360.jpg, -1, -1, 62.6857798165
+4385399670.jpg, -1, -1, 50.4294871795
+4389836139.jpg, -1, -1, 37.5357142857
+4390875397.jpg, -1, -1, 56.0252293578
+4391517237.jpg, -1, -1, 54.8899082569
+4393459228.jpg, -1, -1, 75.8392857143
+4396087406.jpg, -1, -1, 59.4189189189
+4396643142.jpg, -1, -1, 72.925
+44009500.jpg, -1, -1, 5.5
+4401185827.jpg, -1, -1, 46.0953389831
+440297710.jpg, -1, -1, 57.5277777778
+4410900135.jpg, -1, -1, 25.8243243243
+4422726404.jpg, -1, -1, 68.1049107143
+4423499380.jpg, -1, -1, 59.4491525424
+4424481742.jpg, -1, -1, 73.0641025641
+4429610219.jpg, -1, -1, 74.2628205128
+4433497740.jpg, -1, -1, 60.7212389381
+4434357623.jpg, -1, -1, 71.5444915254
+4436568659.jpg, -1, -1, 67.0705128205
+4443705808.jpg, -1, -1, 34.6513157895
+4456369143.jpg, -1, -1, 62.3855932203
+4464186761.jpg, -1, -1, 77.725
+4468108530.jpg, -1, -1, 60.695754717
+4472735511.jpg, -1, -1, 51.925
+4476029726.jpg, -1, -1, 67.2972972973
+4483042022.jpg, -1, -1, 65.7364864865
+448305008.jpg, -1, -1, 66.3067226891
+4484527101.jpg, -1, -1, 66.325
+4491574274.jpg, -1, -1, 71.3421052632
+4492079346.jpg, -1, -1, 60.869266055
+4492298311.jpg, -1, -1, 67.3466386555
+4502409601.jpg, -1, -1, 24.6286407767
+4504786016.jpg, -1, -1, 45.3163716814
+4509028861.jpg, -1, -1, 54.0044247788
+4511318356.jpg, -1, -1, 72.1262135922
+4513142635.jpg, -1, -1, 42.3973214286
+4518358107.jpg, -1, -1, 39.8142857143
+4519546786.jpg, -1, -1, 59.3956521739
+4526077589.jpg, -1, -1, 65.1974789916
+4526456137.jpg, -1, -1, 72.5
+45323583.jpg, -1, -1, 10.0256410256
+4538412115.jpg, -1, -1, 17.7079831933
+4542866851.jpg, -1, -1, 45.2563559322
+4551121129.jpg, -1, -1, 80.9173913043
+4562998069.jpg, -1, -1, 72.1944444444
+456472871.jpg, -1, -1, 62.5803571429
+4567181047.jpg, -1, -1, 69.6490825688
+4567762490.jpg, -1, -1, 73.2756410256
+4568018818.jpg, -1, -1, 73.0127118644
+4570575045.jpg, -1, -1, 57.2941176471
+4580698398.jpg, -1, -1, 34.0711206897
+4583413420.jpg, -1, -1, 66.9243119266
+4589538888.jpg, -1, -1, 44.8
+4597162839.jpg, -1, -1, 74.7297297297
+460024423.jpg, -1, -1, 71.95
+4605589887.jpg, -1, -1, 62.5630252101
+4611014636.jpg, -1, -1, 73.0452586207
+4614592470.jpg, -1, -1, 35.2110091743
+4615199637.jpg, -1, -1, 35.1565420561
+4617747913.jpg, -1, -1, 48.0691964286
+4619633197.jpg, -1, -1, 77.4054054054
+4620749659.jpg, -1, -1, 69.1887755102
+4626914740.jpg, -1, -1, 69.7991071429
+4631127472.jpg, -1, -1, 45.504587156
+463525569.jpg, -1, -1, 69.0435779817
+4638688609.jpg, -1, -1, 72.6155462185
+4644913731.jpg, -1, -1, 58.68125
+4656808355.jpg, -1, -1, 58.1043478261
+4664658605.jpg, -1, -1, 61.3677884615
+4668889560.jpg, -1, -1, 66.5619469027
+4670475171.jpg, -1, -1, 78.1617647059
+4671787703.jpg, -1, -1, 63.4642857143
+4679147725.jpg, -1, -1, 70.0200892857
+4680082599.jpg, -1, -1, 77.6592920354
+4681329966.jpg, -1, -1, 72.5
+468518360.jpg, -1, -1, 64.880733945
+468764619.jpg, -1, -1, 58.6065217391
+4689181626.jpg, -1, -1, 49.7791262136
+4696239138.jpg, -1, -1, 68.9411764706
+4699352200.jpg, -1, -1, 71.9345794393
+4699474732.jpg, -1, -1, 53.3004587156
+4704859608.jpg, -1, -1, 34.7279411765
+4711101546.jpg, -1, -1, 61.5504587156
+4719295519.jpg, -1, -1, 49.3108108108
+4727744330.jpg, -1, -1, 74.2012711864
+4728646090.jpg, -1, -1, 21.296460177
+4731015701.jpg, -1, -1, 54.4078947368
+4731330377.jpg, -1, -1, 51.7375
+4732322374.jpg, -1, -1, 75.8445945946
+4733189524.jpg, -1, -1, 53.7435897436
+473881316.jpg, -1, -1, 16.984375
+4748697171.jpg, -1, -1, 41.2794117647
+4764540278.jpg, -1, -1, 65.4618644068
+4765706226.jpg, -1, -1, 72.0887850467
+4770169005.jpg, -1, -1, 70.3147321429
+4770991170.jpg, -1, -1, 60.925
+4781142781.jpg, -1, -1, 56.1631355932
+4792717107.jpg, -1, -1, 70.4816513761
+4794011269.jpg, -1, -1, 70.2860169492
+4794780770.jpg, -1, -1, 62.1758474576
+4801184295.jpg, -1, -1, 72.7972972973
+4804088480.jpg, -1, -1, 66.2216981132
+4806281586.jpg, -1, -1, 57.5485436893
+4815901200.jpg, -1, -1, 81.0818965517
+4822559131.jpg, -1, -1, 51.9347826087
+4822965011.jpg, -1, -1, 63.3728448276
+4831070004.jpg, -1, -1, 11.9294871795
+4835256174.jpg, -1, -1, 79.7433035714
+4835865863.jpg, -1, -1, 56.6875
+4841257153.jpg, -1, -1, 54.6949152542
+4845997437.jpg, -1, -1, 49.6518691589
+4848345613.jpg, -1, -1, 75.7435897436
+4851066149.jpg, -1, -1, 34.6875
+485174290.jpg, -1, -1, 67.2972972973
+4852617544.jpg, -1, -1, 72.2063106796
+4858527470.jpg, -1, -1, 48.50625
+4862383971.jpg, -1, -1, 65.7547169811
+4863093063.jpg, -1, -1, 41.3173913043
+4867994945.jpg, -1, -1, 21.0256410256
+4871847706.jpg, -1, -1, 72.5743243243
+4873047491.jpg, -1, -1, 74.5448717949
+4875935356.jpg, -1, -1, 58.6829268293
+4875993491.jpg, -1, -1, 63.4426605505
+4885160336.jpg, -1, -1, 72.1038135593
+4885928195.jpg, -1, -1, 52.2217391304
+4886153819.jpg, -1, -1, 62.95
+4890155523.jpg, -1, -1, 74.7582781457
+4891468275.jpg, -1, -1, 77.7027027027
+489975423.jpg, -1, -1, 71.675
+4903384785.jpg, -1, -1, 63.2094594595
+4903418699.jpg, -1, -1, 49.33125
+4922860865.jpg, -1, -1, 47.0379464286
+4922892935.jpg, -1, -1, 74.3411016949
+4923537063.jpg, -1, -1, 56.3279816514
+4925153617.jpg, -1, -1, 69.322815534
+4925167261.jpg, -1, -1, 60.968487395
+4926787829.jpg, -1, -1, 55.1939655172
+4933863195.jpg, -1, -1, 48.7162162162
+4938188075.jpg, -1, -1, 47.4418103448
+4941401132.jpg, -1, -1, 76.5758928571
+4947500856.jpg, -1, -1, 73.3276699029
+4947572555.jpg, -1, -1, 69.9027777778
+4948021193.jpg, -1, -1, 62.6652542373
+4948125167.jpg, -1, -1, 63.90625
+4955052105.jpg, -1, -1, 61.6195652174
+4955674278.jpg, -1, -1, 59.1299342105
+4962117985.jpg, -1, -1, 32.3919491525
+4967412135.jpg, -1, -1, 67.5789473684
+4969822064.jpg, -1, -1, 56.1766055046
+497163401.jpg, -1, -1, 73.7118644068
+4972553952.jpg, -1, -1, 65.5878378378
+4980829250.jpg, -1, -1, 26.7288135593
+4981035833.jpg, -1, -1, 69.925
+4981184708.jpg, -1, -1, 55.5887850467
+4982428461.jpg, -1, -1, 65.8564356436
+4985465789.jpg, -1, -1, 61.6033834586
+4986076377.jpg, -1, -1, 69.55
+4986510349.jpg, -1, -1, 21.4152542373
+4986682017.jpg, -1, -1, 13.5677570093
+49874259.jpg, -1, -1, 62.4318181818
+4995119944.jpg, -1, -1, 73.6155660377
+5001805928.jpg, -1, -1, 40.1689189189
+5001866078.jpg, -1, -1, 65.8051724138
+5005558786.jpg, -1, -1, 71.7142857143
+5005908077.jpg, -1, -1, 64.063559322
+5014526130.jpg, -1, -1, 69.9154135338
+5016657118.jpg, -1, -1, 49.4454887218
+5017267795.jpg, -1, -1, 61.6195652174
+5017618497.jpg, -1, -1, 65.1822033898
+502182808.jpg, -1, -1, 71.7683486239
+5027882580.jpg, -1, -1, 78.6513157895
+5028389463.jpg, -1, -1, 48.8808411215
+503291895.jpg, -1, -1, 64.391025641
+5035645781.jpg, -1, -1, 57.7875
+5036262446.jpg, -1, -1, 68.2336448598
+5037503230.jpg, -1, -1, 61.9390756303
+5038604558.jpg, -1, -1, 78.222972973
+5041593593.jpg, -1, -1, 47.8986486486
+5044457387.jpg, -1, -1, 64.294
+5050399849.jpg, -1, -1, 54.5935114504
+5051600612.jpg, -1, -1, 63.08125
+5055213857.jpg, -1, -1, 48.8194444444
+5062110567.jpg, -1, -1, 48.6184210526
+5064176245.jpg, -1, -1, 75.1756756757
+5073500906.jpg, -1, -1, 58.825
+5075275205.jpg, -1, -1, 76.00625
+5077869968.jpg, -1, -1, 42.8289473684
+5079448176.jpg, -1, -1, 68.8177966102
+5082407583.jpg, -1, -1, 56.6592465753
+5090174589.jpg, -1, -1, 30.49375
+5090549164.jpg, -1, -1, 66.3599137931
+5090803058.jpg, -1, -1, 54.3303571429
+5094262203.jpg, -1, -1, 68.7331932773
+5097068931.jpg, -1, -1, 82.7361111111
+5101108569.jpg, -1, -1, 45.1195652174
+5103528737.jpg, -1, -1, 33.1897321429
+5115371206.jpg, -1, -1, 57.3119266055
+5120849681.jpg, -1, -1, 35.9642857143
+5129172.jpg, -1, -1, 18.5399159664
+5135958105.jpg, -1, -1, 47.6995412844
+5137221601.jpg, -1, -1, 42.5320512821
+5140673192.jpg, -1, -1, 57.561440678
+5142063764.jpg, -1, -1, 40.8181818182
+514498047.jpg, -1, -1, 56.2455357143
+5146611755.jpg, -1, -1, 48.6495327103
+5150999709.jpg, -1, -1, 41.5567226891
+5153880006.jpg, -1, -1, 74.9357142857
+5157059497.jpg, -1, -1, 61.8961864407
+5164262533.jpg, -1, -1, 64.3716814159
+5166185394.jpg, -1, -1, 62.8
+5166352802.jpg, -1, -1, 53.8392857143
+5166736586.jpg, -1, -1, 79.2692307692
+5167059215.jpg, -1, -1, 75.7798165138
+5170785709.jpg, -1, -1, 65.5212264151
+5172565267.jpg, -1, -1, 41.6858974359
+5172812012.jpg, -1, -1, 56.61875
+5176681750.jpg, -1, -1, 67.1386554622
+5179986164.jpg, -1, -1, 50.9805825243
+5183223776.jpg, -1, -1, 69.3771186441
+5184506144.jpg, -1, -1, 81.5588235294
+5185255513.jpg, -1, -1, 72.8125
+5189289022.jpg, -1, -1, 77.3486842105
+5193757906.jpg, -1, -1, 59.575
+5195745450.jpg, -1, -1, 56.4036697248
+5197644731.jpg, -1, -1, 56.570754717
+5209757780.jpg, -1, -1, 58.1207627119
+5216419230.jpg, -1, -1, 71.8310810811
+5216870847.jpg, -1, -1, 52.7434210526
+5217550500.jpg, -1, -1, 29.7193396226
+5221875216.jpg, -1, -1, 47.9830508475
+5223615087.jpg, -1, -1, 74.4676724138
+5233564371.jpg, -1, -1, 60.0897435897
+5236886996.jpg, -1, -1, 73.0401785714
+5237114997.jpg, -1, -1, 31.8508403361
+5241841036.jpg, -1, -1, 77.51875
+5249052587.jpg, -1, -1, 57.6313559322
+5251948512.jpg, -1, -1, 71.3860759494
+5252606224.jpg, -1, -1, 63.3155339806
+5258392332.jpg, -1, -1, 67.675
+526016794.jpg, -1, -1, 45.1703539823
+5261364718.jpg, -1, -1, 61.6527777778
+526555735.jpg, -1, -1, 50.1972477064
+5266542779.jpg, -1, -1, 64.502293578
+526762082.jpg, -1, -1, 67.8319327731
+5269040313.jpg, -1, -1, 12.2433628319
+5269333343.jpg, -1, -1, 79.15
+5272496469.jpg, -1, -1, 34.9491150442
+5276700957.jpg, -1, -1, 42.1800847458
+5286481593.jpg, -1, -1, 65.6004464286
+5288273177.jpg, -1, -1, 58.3608490566
+528858440.jpg, -1, -1, 36.5826086957
+5292802931.jpg, -1, -1, 40.4077669903
+5293214322.jpg, -1, -1, 72.0609243697
+5297588508.jpg, -1, -1, 53.1283185841
+5298103951.jpg, -1, -1, 40.9121621622
+5305142351.jpg, -1, -1, 71.9646017699
+5305438208.jpg, -1, -1, 45.8855932203
+5308844734.jpg, -1, -1, 74.8816964286
+5309442831.jpg, -1, -1, 73.5131578947
+5309538408.jpg, -1, -1, 57.7094594595
+5312388504.jpg, -1, -1, 68.7633587786
+5317045814.jpg, -1, -1, 66.621559633
+5320512069.jpg, -1, -1, 46.6594827586
+5326008096.jpg, -1, -1, 64.5970873786
+5327737580.jpg, -1, -1, 52.3982300885
+5328075993.jpg, -1, -1, 66.4054054054
+5342778275.jpg, -1, -1, 55.4646017699
+5343739846.jpg, -1, -1, 46.16875
+5352670374.jpg, -1, -1, 66.691588785
+535787632.jpg, -1, -1, 38.5995575221
+5358352176.jpg, -1, -1, 40.6605504587
+5358640896.jpg, -1, -1, 68.275
+5359127660.jpg, -1, -1, 67.7891304348
+5360175600.jpg, -1, -1, 75.6183035714
+5362095529.jpg, -1, -1, 75.3986486486
+5363150332.jpg, -1, -1, 55.203271028
+5364427287.jpg, -1, -1, 69.2564102564
+5365003071.jpg, -1, -1, 67.20625
+5370538873.jpg, -1, -1, 50.1381578947
+5371908851.jpg, -1, -1, 66.2699115044
+537232636.jpg, -1, -1, 40.2960526316
+5375536055.jpg, -1, -1, 67.9871794872
+5376223041.jpg, -1, -1, 68.0405405405
+5382638405.jpg, -1, -1, 44.858974359
+5399327452.jpg, -1, -1, 63.325
+5407907045.jpg, -1, -1, 48.1637168142
+5407931669.jpg, -1, -1, 73.6308411215
+5414089487.jpg, -1, -1, 80.4146341463
+5414222209.jpg, -1, -1, 77.9740566038
+5419570020.jpg, -1, -1, 66.7053571429
+5420375203.jpg, -1, -1, 74.2382075472
+5424903451.jpg, -1, -1, 68.5218446602
+5430352868.jpg, -1, -1, 71.0884955752
+5430572465.jpg, -1, -1, 69.9203539823
+5433364879.jpg, -1, -1, 30.256302521
+5434982423.jpg, -1, -1, 38.0684931507
+5436442410.jpg, -1, -1, 70.0394736842
+5437565452.jpg, -1, -1, 62.7635135135
+5457535055.jpg, -1, -1, 64.3986486486
+5459749774.jpg, -1, -1, 49.7556390977
+5460394455.jpg, -1, -1, 75.765625
+5464519026.jpg, -1, -1, 48.55
+5471871046.jpg, -1, -1, 35.0951327434
+5473035889.jpg, -1, -1, 73.1346153846
+5482786130.jpg, -1, -1, 76.5413043478
+5483273817.jpg, -1, -1, 56.3407079646
+5483697378.jpg, -1, -1, 53.8147321429
+5483904990.jpg, -1, -1, 53.1013513514
+5487766524.jpg, -1, -1, 50.0805084746
+5489335291.jpg, -1, -1, 76.4911504425
+5494503911.jpg, -1, -1, 72.6009174312
+5495773051.jpg, -1, -1, 62.5803571429
+5496213154.jpg, -1, -1, 73.476635514
+5501090896.jpg, -1, -1, 61.7563559322
+5504474324.jpg, -1, -1, 54.2430555556
+5505953336.jpg, -1, -1, 20.2731092437
+5508282751.jpg, -1, -1, 74.3092105263
+5508551413.jpg, -1, -1, 26.3785714286
+5512989538.jpg, -1, -1, 69.3469827586
+5519011400.jpg, -1, -1, 54.14375
+5525531268.jpg, -1, -1, 47.9446902655
+552629141.jpg, -1, -1, 30.5379464286
+5528731780.jpg, -1, -1, 68.8877118644
+5531274580.jpg, -1, -1, 66.4701834862
+5534328979.jpg, -1, -1, 50.3526785714
+5538586634.jpg, -1, -1, 77.9766949153
+5539522093.jpg, -1, -1, 70.6844660194
+5543037671.jpg, -1, -1, 63.7475961538
+5544820783.jpg, -1, -1, 42.8289473684
+5546833927.jpg, -1, -1, 62.1734693878
+5546979604.jpg, -1, -1, 72.8928571429
+5558342101.jpg, -1, -1, 61.8961864407
+5561303733.jpg, -1, -1, 34.8561320755
+5561541116.jpg, -1, -1, 45.3163716814
+5569876905.jpg, -1, -1, 48.4025423729
+5572325411.jpg, -1, -1, 74.5625
+5573505825.jpg, -1, -1, 74.2888349515
+5578143578.jpg, -1, -1, 72.443877551
+5580904626.jpg, -1, -1, 72.8598130841
+5586376319.jpg, -1, -1, 61.7853773585
+5589406824.jpg, -1, -1, 40.6605504587
+5590179803.jpg, -1, -1, 57.475
+5592040758.jpg, -1, -1, 69.1673728814
+5593028853.jpg, -1, -1, 61.8341121495
+5599500585.jpg, -1, -1, 53.8147321429
+5599655175.jpg, -1, -1, 76.3879310345
+560054684.jpg, -1, -1, 21.0573394495
+5602088551.jpg, -1, -1, 67.2115384615
+5602195949.jpg, -1, -1, 71.5833333333
+5602672763.jpg, -1, -1, 61.6375
+5606693622.jpg, -1, -1, 61.5231092437
+5607542951.jpg, -1, -1, 48.6418918919
+5609386026.jpg, -1, -1, 56.7079207921
+5621762674.jpg, -1, -1, 81.320754717
+5625316207.jpg, -1, -1, 64.357421875
+5627315065.jpg, -1, -1, 73.8397435897
+562942809.jpg, -1, -1, 21.2027559055
+5631644954.jpg, -1, -1, 48.0818965517
+5632060967.jpg, -1, -1, 65.4487179487
+5634844318.jpg, -1, -1, 43.9622641509
+5645276165.jpg, -1, -1, 63.5810810811
+5646738672.jpg, -1, -1, 62.1383928571
+5650628470.jpg, -1, -1, 56.2607758621
+5651165439.jpg, -1, -1, 79.9976415094
+5655651548.jpg, -1, -1, 32.3642241379
+5664632199.jpg, -1, -1, 59.5602678571
+5666479112.jpg, -1, -1, 65.685840708
+5669648365.jpg, -1, -1, 59.0527522936
+5673514295.jpg, -1, -1, 70.3277310924
+5680237187.jpg, -1, -1, 79.0066964286
+5680571645.jpg, -1, -1, 31.027027027
+5683736894.jpg, -1, -1, 32.3919491525
+5684894057.jpg, -1, -1, 60.325
+5685822958.jpg, -1, -1, 79.6696428571
+5687490427.jpg, -1, -1, 44.2948717949
+5689072971.jpg, -1, -1, 46.1857798165
+5690961665.jpg, -1, -1, 76.9151376147
+5692339429.jpg, -1, -1, 72.0217391304
+5692933688.jpg, -1, -1, 72.9342105263
+5694228883.jpg, -1, -1, 71.447826087
+5698690184.jpg, -1, -1, 69.4127358491
+5699749992.jpg, -1, -1, 67.5689655172
+5700224418.jpg, -1, -1, 50.8716216216
+5700803236.jpg, -1, -1, 63.9724770642
+5701314350.jpg, -1, -1, 75.5420353982
+5702257196.jpg, -1, -1, 51.506097561
+5706720351.jpg, -1, -1, 30.256302521
+5718858871.jpg, -1, -1, 69.9575471698
+5719653094.jpg, -1, -1, 17.2920168067
+5737344773.jpg, -1, -1, 68.6638655462
+5740846112.jpg, -1, -1, 55.8035714286
+5746784745.jpg, -1, -1, 46.375
+5747013682.jpg, -1, -1, 28.075
+5748221179.jpg, -1, -1, 60.8672566372
+5748620149.jpg, -1, -1, 56.9821428571
+5748929912.jpg, -1, -1, 43.9831932773
+57501165.jpg, -1, -1, 51.8983050847
+5756631011.jpg, -1, -1, 59.1959459459
+5763062464.jpg, -1, -1, 73.7767857143
+5764115414.jpg, -1, -1, 71.7142857143
+5764514076.jpg, -1, -1, 29.0079617834
+5771711136.jpg, -1, -1, 59.9181415929
+5771887382.jpg, -1, -1, 26.0966386555
+5790219836.jpg, -1, -1, 61.2330097087
+5796436033.jpg, -1, -1, 60.74375
+5798200461.jpg, -1, -1, 71.2326086957
+5798775388.jpg, -1, -1, 49.8370535714
+5803914327.jpg, -1, -1, 43.2905405405
+5805954950.jpg, -1, -1, 72.8470873786
+5812785774.jpg, -1, -1, 76.0602678571
+5817880168.jpg, -1, -1, 69.119266055
+582148952.jpg, -1, -1, 69.8727678571
+5827460813.jpg, -1, -1, 60.2881355932
+5830956184.jpg, -1, -1, 63.1136363636
+5835525141.jpg, -1, -1, 71.35
+5836498446.jpg, -1, -1, 67.6292372881
+5838518725.jpg, -1, -1, 65.8108108108
+5842574442.jpg, -1, -1, 80.2311320755
+5842712003.jpg, -1, -1, 70.3584070796
+5844921100.jpg, -1, -1, 68.6923076923
+58449294.jpg, -1, -1, 50.6363636364
+5853259008.jpg, -1, -1, 69.7013274336
+5865132229.jpg, -1, -1, 57.9180672269
+5866763424.jpg, -1, -1, 73.3205645161
+5870955720.jpg, -1, -1, 74.7102803738
+5871068150.jpg, -1, -1, 62.7710084034
+5871593028.jpg, -1, -1, 64.5169902913
+5875925963.jpg, -1, -1, 54.7307692308
+5876722447.jpg, -1, -1, 58.6699029126
+5877192304.jpg, -1, -1, 58.9690265487
+5877272501.jpg, -1, -1, 79.1956521739
+5881856929.jpg, -1, -1, 56.7279411765
+5884906716.jpg, -1, -1, 71.1949152542
+5886086971.jpg, -1, -1, 44.9471153846
+5891588921.jpg, -1, -1, 75.25
+5893652818.jpg, -1, -1, 18.7132352941
+5894875244.jpg, -1, -1, 76.6378504673
+5898928772.jpg, -1, -1, 55.495412844
+5899300375.jpg, -1, -1, 39.925
+5903878980.jpg, -1, -1, 67.6939252336
+5903930781.jpg, -1, -1, 72.7669902913
+5905277829.jpg, -1, -1, 57.0861344538
+5910545670.jpg, -1, -1, 30.390625
+5910732650.jpg, -1, -1, 25.2201834862
+5912436872.jpg, -1, -1, 66.4054054054
+5914413530.jpg, -1, -1, 66.9269911504
+5917296642.jpg, -1, -1, 27.5665137615
+5921291430.jpg, -1, -1, 64.063559322
+5926348208.jpg, -1, -1, 59.1183035714
+5937141958.jpg, -1, -1, 65.1184210526
+5944469044.jpg, -1, -1, 61.0456521739
+5945639421.jpg, -1, -1, 56.171875
+5948378050.jpg, -1, -1, 60.2428571429
+5948782020.jpg, -1, -1, 25.8193277311
+5955821366.jpg, -1, -1, 63.6029411765
+5956532834.jpg, -1, -1, 19.7706422018
+5958527369.jpg, -1, -1, 76.1588983051
+5959274849.jpg, -1, -1, 47.26875
+5960283649.jpg, -1, -1, 43.7625
+5965181934.jpg, -1, -1, 61.5257009346
+5966712198.jpg, -1, -1, 58.8922413793
+5966778623.jpg, -1, -1, 36.25
+5968558261.jpg, -1, -1, 73.9473684211
+5969471767.jpg, -1, -1, 48.9105504587
+5971762123.jpg, -1, -1, 71.8616071429
+5977841815.jpg, -1, -1, 82.5673913043
+5985353334.jpg, -1, -1, 64.502293578
+5987384542.jpg, -1, -1, 82.8188073394
+5987658762.jpg, -1, -1, 69.0625
+5996266717.jpg, -1, -1, 72.198630137
+5997968998.jpg, -1, -1, 63.325
+6001258616.jpg, -1, -1, 64.5691964286
+6001836316.jpg, -1, -1, 44.6764705882
+6004179446.jpg, -1, -1, 58.15
+6004662053.jpg, -1, -1, 71.9934210526
+6011321842.jpg, -1, -1, 62.2
+6012494248.jpg, -1, -1, 32.9882075472
+6014996044.jpg, -1, -1, 21.947826087
+6015032576.jpg, -1, -1, 44.777027027
+6018292739.jpg, -1, -1, 82.825
+6027079771.jpg, -1, -1, 35.6205357143
+6031428768.jpg, -1, -1, 76.7697368421
+6031659162.jpg, -1, -1, 70.0663716814
+6038209414.jpg, -1, -1, 67.7638888889
+6050570749.jpg, -1, -1, 45.5909090909
+6051776132.jpg, -1, -1, 77.6592920354
+6052522172.jpg, -1, -1, 65.4875
+6063249330.jpg, -1, -1, 58.825
+6063545051.jpg, -1, -1, 46.5959821429
+6064738730.jpg, -1, -1, 71.9338235294
+6065704403.jpg, -1, -1, 67.7432432432
+6072006649.jpg, -1, -1, 67.6823308271
+6073506056.jpg, -1, -1, 71.2648305085
+6077061574.jpg, -1, -1, 68.468220339
+6078472527.jpg, -1, -1, 61.3053097345
+6084731334.jpg, -1, -1, 38.9197247706
+6092350487.jpg, -1, -1, 70.5675675676
+6093133362.jpg, -1, -1, 51.9935344828
+6098120607.jpg, -1, -1, 23.7989864865
+6099389158.jpg, -1, -1, 36.4605263158
+6113347868.jpg, -1, -1, 51.7584745763
+6114847662.jpg, -1, -1, 64.3685344828
+6115396985.jpg, -1, -1, 70.7087155963
+6116188135.jpg, -1, -1, 72.8728813559
+6118735934.jpg, -1, -1, 76.397826087
+6118760299.jpg, -1, -1, 77.8962264151
+6124044226.jpg, -1, -1, 71.9797297297
+6127748837.jpg, -1, -1, 67.9491150442
+6131345309.jpg, -1, -1, 69.5108695652
+6138363884.jpg, -1, -1, 72.8928571429
+6142185721.jpg, -1, -1, 39.6009174312
+6143698715.jpg, -1, -1, 33.6349557522
+6144690260.jpg, -1, -1, 76.4399038462
+6144702078.jpg, -1, -1, 54.0357142857
+6146146023.jpg, -1, -1, 80.7739130435
+6146595160.jpg, -1, -1, 71.6261682243
+6147638271.jpg, -1, -1, 76.5486111111
+6149089572.jpg, -1, -1, 73.3517699115
+6153962968.jpg, -1, -1, 40.6605504587
+6156825791.jpg, -1, -1, 66.0388349515
+6164110615.jpg, -1, -1, 56.3826086957
+6166626410.jpg, -1, -1, 48.9353448276
+6170589548.jpg, -1, -1, 64.4348739496
+6172964730.jpg, -1, -1, 23.8224299065
+6173003449.jpg, -1, -1, 72.6009174312
+61745002.jpg, -1, -1, 47.7727272727
+6188110673.jpg, -1, -1, 75.1026785714
+6190085634.jpg, -1, -1, 51.8865546218
+6190639250.jpg, -1, -1, 72.5827067669
+6192537618.jpg, -1, -1, 71.9122137405
+6199370121.jpg, -1, -1, 76.4695652174
+6210494281.jpg, -1, -1, 59.7075892857
+6216457478.jpg, -1, -1, 64.9195652174
+6218336714.jpg, -1, -1, 75.1794871795
+6219475255.jpg, -1, -1, 61.1808035714
+6220758013.jpg, -1, -1, 67.6629464286
+6224309834.jpg, -1, -1, 23.3055555556
+6226065978.jpg, -1, -1, 65.8851351351
+6232031392.jpg, -1, -1, 71.8241525424
+6237907286.jpg, -1, -1, 67.1473214286
+62458506.jpg, -1, -1, 64.5320512821
+6248527465.jpg, -1, -1, 75.6391509434
+6251598886.jpg, -1, -1, 53.3727678571
+6251795586.jpg, -1, -1, 73.4878640777
+6252497829.jpg, -1, -1, 57.8660714286
+6258118545.jpg, -1, -1, 36.5826086957
+6264153932.jpg, -1, -1, 40.6605504587
+6267817145.jpg, -1, -1, 64.975
+6268176546.jpg, -1, -1, 75.8739495798
+6274950285.jpg, -1, -1, 76.675
+6276334186.jpg, -1, -1, 35.44375
+6280923250.jpg, -1, -1, 69.7737068966
+6283627097.jpg, -1, -1, 57.8247663551
+6285601097.jpg, -1, -1, 73.6933962264
+6290479072.jpg, -1, -1, 64.2919847328
+6293629106.jpg, -1, -1, 44.8183962264
+6294913930.jpg, -1, -1, 58.75
+6298978484.jpg, -1, -1, 59.8096330275
+6303893419.jpg, -1, -1, 56.570754717
+6304784314.jpg, -1, -1, 28.3513513514
+6305224063.jpg, -1, -1, 24.8891304348
+6305573914.jpg, -1, -1, 53.7715517241
+6310649828.jpg, -1, -1, 18.7522321429
+6317232957.jpg, -1, -1, 61.5504587156
+6318820907.jpg, -1, -1, 61.8819444444
+6319935355.jpg, -1, -1, 82.1741071429
+6321752444.jpg, -1, -1, 57.4864864865
+6325240412.jpg, -1, -1, 79.9343220339
+6326811967.jpg, -1, -1, 64.6025641026
+6326902733.jpg, -1, -1, 66.2633928571
+6332883956.jpg, -1, -1, 16.9
+6333112690.jpg, -1, -1, 67.502173913
+6340619552.jpg, -1, -1, 30.4324324324
+6344881273.jpg, -1, -1, 54.925
+6391074279.jpg, -1, -1, 76.6973684211
+6391919133.jpg, -1, -1, 64.4446902655
+6403149971.jpg, -1, -1, 35.3231132075
+6408708033.jpg, -1, -1, 47.7008928571
+6411944553.jpg, -1, -1, 56.9977876106
+641595598.jpg, -1, -1, 74.747826087
+6419047435.jpg, -1, -1, 67.8009708738
+6422846525.jpg, -1, -1, 71.6169724771
+6433692937.jpg, -1, -1, 45.7742718447
+6435747701.jpg, -1, -1, 74.3173913043
+6441383857.jpg, -1, -1, 74.1603773585
+6443396365.jpg, -1, -1, 56.0233050847
+6443711585.jpg, -1, -1, 45.4459459459
+6445980191.jpg, -1, -1, 63.8109243697
+6453185751.jpg, -1, -1, 60.5525210084
+6453831461.jpg, -1, -1, 38.0871559633
+6458056243.jpg, -1, -1, 42.6880530973
+6459811757.jpg, -1, -1, 68.2913043478
+6467546565.jpg, -1, -1, 29.7710084034
+6470955539.jpg, -1, -1, 64.4316037736
+6471980333.jpg, -1, -1, 59.7288135593
+6481674971.jpg, -1, -1, 53.425
+6486906969.jpg, -1, -1, 38.2385321101
+6489144861.jpg, -1, -1, 66.9256756757
+6495386031.jpg, -1, -1, 68.9576271186
+6496266341.jpg, -1, -1, 35.8982300885
+6497142149.jpg, -1, -1, 63.0148305085
+6511236583.jpg, -1, -1, 68.4558823529
+6512617013.jpg, -1, -1, 80.26875
+6517362491.jpg, -1, -1, 61.9504310345
+6522952765.jpg, -1, -1, 61.626146789
+6523676619.jpg, -1, -1, 56.4036697248
+6530161213.jpg, -1, -1, 54.7648305085
+6543602941.jpg, -1, -1, 55.3185840708
+6563196289.jpg, -1, -1, 52.6554054054
+6563980965.jpg, -1, -1, 81.1625
+6564941395.jpg, -1, -1, 72.3296460177
+6568238645.jpg, -1, -1, 56.9247787611
+6577277115.jpg, -1, -1, 62.5485611511
+6578628541.jpg, -1, -1, 55.768907563
+6579889887.jpg, -1, -1, 64.1334745763
+6585714605.jpg, -1, -1, 80.40625
+6595488751.jpg, -1, -1, 64.343220339
+6596085063.jpg, -1, -1, 62.1111111111
+6596306.jpg, -1, -1, 40.6282051282
+6609931137.jpg, -1, -1, 67.2115384615
+6613235741.jpg, -1, -1, 42.3308823529
+6621488051.jpg, -1, -1, 66.5840336134
+6622717529.jpg, -1, -1, 74.5387931034
+6632573597.jpg, -1, -1, 53.125
+6633974935.jpg, -1, -1, 74.2794117647
+6638453859.jpg, -1, -1, 75.85
+6645063001.jpg, -1, -1, 43.3686440678
+6648576397.jpg, -1, -1, 40.1470588235
+6658108283.jpg, -1, -1, 26.9695652174
+6658947687.jpg, -1, -1, 47.6995412844
+6661741471.jpg, -1, -1, 19.8771186441
+6669862645.jpg, -1, -1, 58.516509434
+6679373485.jpg, -1, -1, 42.6625
+6681716765.jpg, -1, -1, 69.9870689655
+6688906197.jpg, -1, -1, 59.9065420561
+6692795337.jpg, -1, -1, 72.3834745763
+6694664547.jpg, -1, -1, 66.4602803738
+6694669509.jpg, -1, -1, 64.6869158879
+6698669369.jpg, -1, -1, 79.9130434783
+6708806483.jpg, -1, -1, 69.1491596639
+6710865793.jpg, -1, -1, 63.8325892857
+6727661759.jpg, -1, -1, 60.4138655462
+6728176335.jpg, -1, -1, 66.9306722689
+6728215257.jpg, -1, -1, 64.9724576271
+6731073163.jpg, -1, -1, 65.6602564103
+6733542819.jpg, -1, -1, 60.0758928571
+6739568529.jpg, -1, -1, 57.9789719626
+6741097253.jpg, -1, -1, 57.8333333333
+6746835381.jpg, -1, -1, 79.8602941176
+6748287295.jpg, -1, -1, 66.3653846154
+6771336533.jpg, -1, -1, 31.1272321429
+6776603937.jpg, -1, -1, 66.7401574803
+6791461596.jpg, -1, -1, 60.2058823529
+6793483796.jpg, -1, -1, 67.0
+6793575356.jpg, -1, -1, 57.7094594595
+6796759928.jpg, -1, -1, 75.8082706767
+6798563238.jpg, -1, -1, 52.1019417476
+6803183748.jpg, -1, -1, 58.4553571429
+6808449405.jpg, -1, -1, 67.8481308411
+6809126683.jpg, -1, -1, 46.2614678899
+6811037194.jpg, -1, -1, 38.1621621622
+6813481975.jpg, -1, -1, 68.4637096774
+6818984013.jpg, -1, -1, 49.5911016949
+6824806522.jpg, -1, -1, 30.9929906542
+6829439085.jpg, -1, -1, 55.5769230769
+6830281130.jpg, -1, -1, 76.1108695652
+6832446699.jpg, -1, -1, 51.506097561
+6833109332.jpg, -1, -1, 70.8076923077
+6834514891.jpg, -1, -1, 28.8081896552
+6836874905.jpg, -1, -1, 68.2954545455
+6837228368.jpg, -1, -1, 69.5169491525
+6837662382.jpg, -1, -1, 71.5265486726
+6842750557.jpg, -1, -1, 29.4166666667
+6843523713.jpg, -1, -1, 56.2894736842
+6843525285.jpg, -1, -1, 71.534351145
+6845847924.jpg, -1, -1, 42.6767241379
+6854228547.jpg, -1, -1, 58.0448717949
+6860404399.jpg, -1, -1, 41.5625
+6864569694.jpg, -1, -1, 65.8851351351
+6865774865.jpg, -1, -1, 66.51875
+6869450975.jpg, -1, -1, 47.608974359
+6869855884.jpg, -1, -1, 35.1402439024
+6870596537.jpg, -1, -1, 56.9084821429
+6880980024.jpg, -1, -1, 29.7163461538
+6881321836.jpg, -1, -1, 18.6866438356
+6882017804.jpg, -1, -1, 46.9642857143
+6890378940.jpg, -1, -1, 75.775
+6890677809.jpg, -1, -1, 80.05
+6891596982.jpg, -1, -1, 64.6311881188
+6891599340.jpg, -1, -1, 63.2354368932
+6892058753.jpg, -1, -1, 68.85625
+6895353221.jpg, -1, -1, 48.5227272727
+6897314758.jpg, -1, -1, 73.5091743119
+6902342114.jpg, -1, -1, 47.8014018692
+6902444479.jpg, -1, -1, 68.8581081081
+6904675638.jpg, -1, -1, 43.0210280374
+6905808386.jpg, -1, -1, 61.3233944954
+6908900076.jpg, -1, -1, 80.3881578947
+6911586648.jpg, -1, -1, 29.8013392857
+6920118996.jpg, -1, -1, 64.9375
+6927641976.jpg, -1, -1, 42.3973214286
+6929188980.jpg, -1, -1, 70.8648648649
+6929191226.jpg, -1, -1, 48.0691964286
+6930075142.jpg, -1, -1, 51.2208737864
+6931269556.jpg, -1, -1, 60.8298319328
+6933526580.jpg, -1, -1, 65.0256410256
+6934705817.jpg, -1, -1, 52.9957983193
+6935047173.jpg, -1, -1, 74.2711864407
+6937606166.jpg, -1, -1, 74.7434210526
+6938468092.jpg, -1, -1, 63.15
+6939374638.jpg, -1, -1, 60.5375
+6940202479.jpg, -1, -1, 65.1495327103
+6940429128.jpg, -1, -1, 40.225
+6941715715.jpg, -1, -1, 64.7367256637
+6942418274.jpg, -1, -1, 54.3823529412
+6942738260.jpg, -1, -1, 64.3947368421
+6945167278.jpg, -1, -1, 70.5902777778
+6945246052.jpg, -1, -1, 77.535472973
+6946044735.jpg, -1, -1, 43.4285714286
+6947747616.jpg, -1, -1, 62.875
+6949001721.jpg, -1, -1, 50.70625
+6954316003.jpg, -1, -1, 59.7899159664
+6954547933.jpg, -1, -1, 37.1032110092
+6955490571.jpg, -1, -1, 72.8598130841
+6962031247.jpg, -1, -1, 39.1658878505
+6963387147.jpg, -1, -1, 56.171875
+6966797840.jpg, -1, -1, 39.3971962617
+6974731766.jpg, -1, -1, 35.3995535714
+6976429442.jpg, -1, -1, 36.8004587156
+6976437867.jpg, -1, -1, 71.6144067797
+6985846441.jpg, -1, -1, 61.4471153846
+6994313445.jpg, -1, -1, 79.046460177
+6996264658.jpg, -1, -1, 62.7654867257
+6998687310.jpg, -1, -1, 56.2115384615
+6998702838.jpg, -1, -1, 63.8325892857
+7001480005.jpg, -1, -1, 59.6418918919
+7003526811.jpg, -1, -1, 68.6351351351
+7005136305.jpg, -1, -1, 78.7544247788
+7005672625.jpg, -1, -1, 66.775
+7006191446.jpg, -1, -1, 53.5762711864
+7013076503.jpg, -1, -1, 64.375
+7013643069.jpg, -1, -1, 44.6627358491
+7024488001.jpg, -1, -1, 78.3163716814
+7025630527.jpg, -1, -1, 62.4734513274
+70272233.jpg, -1, -1, 15.8361344538
+7033122257.jpg, -1, -1, 55.3254716981
+7033769199.jpg, -1, -1, 54.1504424779
+7036262065.jpg, -1, -1, 71.7763157895
+7043681927.jpg, -1, -1, 70.2715517241
+7046294773.jpg, -1, -1, 72.9665178571
+7051563883.jpg, -1, -1, 57.8891304348
+7052391619.jpg, -1, -1, 76.8175
+7054584681.jpg, -1, -1, 65.6621621622
+7063436849.jpg, -1, -1, 60.1116504854
+706409343.jpg, -1, -1, 72.0887850467
+7066508685.jpg, -1, -1, 53.6788990826
+7066539033.jpg, -1, -1, 71.2307692308
+7078293355.jpg, -1, -1, 79.4102564103
+7082408177.jpg, -1, -1, 49.9701834862
+7096952113.jpg, -1, -1, 62.875
+7110105161.jpg, -1, -1, 47.0986842105
+7113461019.jpg, -1, -1, 51.5076335878
+7119860187.jpg, -1, -1, 70.7234513274
+7131772191.jpg, -1, -1, 79.5642201835
+7136073057.jpg, -1, -1, 62.1689189189
+7144832397.jpg, -1, -1, 68.6351351351
+71458295.jpg, -1, -1, 62.9789915966
+7149994451.jpg, -1, -1, 62.8391304348
+7158901982.jpg, -1, -1, 62.7654867257
+7162678728.jpg, -1, -1, 63.5297619048
+7162855925.jpg, -1, -1, 11.8552631579
+7167671368.jpg, -1, -1, 63.8802521008
+7168526408.jpg, -1, -1, 64.7043478261
+7171867098.jpg, -1, -1, 68.6061946903
+7172192793.jpg, -1, -1, 51.8983050847
+7176672187.jpg, -1, -1, 59.3392857143
+7178199774.jpg, -1, -1, 81.4932432432
+7181400642.jpg, -1, -1, 41.65
+7189264298.jpg, -1, -1, 79.8256302521
+7189498789.jpg, -1, -1, 64.1995412844
+7189997672.jpg, -1, -1, 69.625
+7191194360.jpg, -1, -1, 47.7256637168
+7192279956.jpg, -1, -1, 63.2094594595
+7192785398.jpg, -1, -1, 56.597826087
+7217797182.jpg, -1, -1, 63.3419811321
+7219295732.jpg, -1, -1, 44.125
+7228274052.jpg, -1, -1, 72.55
+7232368208.jpg, -1, -1, 62.6924778761
+7238368946.jpg, -1, -1, 36.8875
+7264923790.jpg, -1, -1, 77.1152173913
+7268981578.jpg, -1, -1, 36.3004807692
+7270764894.jpg, -1, -1, 70.2004310345
+7276446104.jpg, -1, -1, 70.3141025641
+7280314854.jpg, -1, -1, 54.8459821429
+7283094026.jpg, -1, -1, 78.34375
+7286003056.jpg, -1, -1, 70.9285714286
+7291000146.jpg, -1, -1, 66.237394958
+7292878318.jpg, -1, -1, 61.1381578947
+7296337218.jpg, -1, -1, 65.5
+7312266068.jpg, -1, -1, 53.31875
+7312752876.jpg, -1, -1, 50.5693277311
+7316131560.jpg, -1, -1, 71.9646017699
+7316593310.jpg, -1, -1, 74.8305084746
+7328195954.jpg, -1, -1, 72.1652173913
+7331539774.jpg, -1, -1, 74.4539473684
+7334314036.jpg, -1, -1, 75.25
+7351425672.jpg, -1, -1, 76.1119402985
+7355344734.jpg, -1, -1, 76.2899159664
+7355708808.jpg, -1, -1, 67.2920353982
+73558937.jpg, -1, -1, 56.1375
+7358286276.jpg, -1, -1, 7.05
+7359378256.jpg, -1, -1, 71.3805309735
+7360550332.jpg, -1, -1, 72.5763888889
+7383311936.jpg, -1, -1, 37.0436893204
+7387717156.jpg, -1, -1, 39.268907563
+7391334516.jpg, -1, -1, 68.4864864865
+7391355184.jpg, -1, -1, 77.5064102564
+7392719524.jpg, -1, -1, 76.1142857143
+7393631040.jpg, -1, -1, 75.7435897436
+7417878338.jpg, -1, -1, 80.4869565217
+7425526298.jpg, -1, -1, 43.7831858407
+7445548804.jpg, -1, -1, 54.9655963303
+7460005346.jpg, -1, -1, 62.4166666667
+7460698156.jpg, -1, -1, 71.5913043478
+7466229858.jpg, -1, -1, 66.7053571429
+7470692282.jpg, -1, -1, 44.05
+7484588674.jpg, -1, -1, 68.1891891892
+7492007796.jpg, -1, -1, 53.0553097345
+7492033480.jpg, -1, -1, 59.3093220339
+7499781424.jpg, -1, -1, 69.797826087
+7504569144.jpg, -1, -1, 70.0847826087
+7510638426.jpg, -1, -1, 54.0675675676
+7538254742.jpg, -1, -1, 54.1830357143
+7551562932.jpg, -1, -1, 40.9633027523
+7553660370.jpg, -1, -1, 56.6701680672
+7557808298.jpg, -1, -1, 75.7610619469
+7559495698.jpg, -1, -1, 19.0945378151
+7561935496.jpg, -1, -1, 73.1111111111
+7575998494.jpg, -1, -1, 68.7368421053
+7577281784.jpg, -1, -1, 67.3618421053
+7582497202.jpg, -1, -1, 65.95
+7586142684.jpg, -1, -1, 64.6337209302
+7590781464.jpg, -1, -1, 74.6444954128
+7591870206.jpg, -1, -1, 60.1602564103
+7592360042.jpg, -1, -1, 64.25
+7613152858.jpg, -1, -1, 73.9241071429
+7617788578.jpg, -1, -1, 62.05
+7622957154.jpg, -1, -1, 75.8846153846
+7625539278.jpg, -1, -1, 35.5137614679
+7636475996.jpg, -1, -1, 23.1766055046
+7645259944.jpg, -1, -1, 40.2051282051
+7651198192.jpg, -1, -1, 63.197826087
+7660049550.jpg, -1, -1, 63.1305309735
+7662529090.jpg, -1, -1, 61.15625
+7672860636.jpg, -1, -1, 56.3701923077
+7673801638.jpg, -1, -1, 80.4336283186
+7674660830.jpg, -1, -1, 68.1323529412
+7678256566.jpg, -1, -1, 78.0924369748
+7690007638.jpg, -1, -1, 58.75
+7692418948.jpg, -1, -1, 58.4630434783
+7697080640.jpg, -1, -1, 66.1610169492
+7697492926.jpg, -1, -1, 65.3366935484
+7703482662.jpg, -1, -1, 70.4903846154
+7704019566.jpg, -1, -1, 72.0376106195
+7712722502.jpg, -1, -1, 74.6760869565
+7720747032.jpg, -1, -1, 78.3716216216
+7728764770.jpg, -1, -1, 72.5514018692
+7731352704.jpg, -1, -1, 76.00625
+7736222330.jpg, -1, -1, 58.3495145631
+7737023436.jpg, -1, -1, 50.0021551724
+7738480598.jpg, -1, -1, 31.3947368421
+7751690228.jpg, -1, -1, 71.6169724771
+7759565946.jpg, -1, -1, 79.0344036697
+7765809112.jpg, -1, -1, 51.4705882353
+7769931466.jpg, -1, -1, 46.9642857143
+7769932398.jpg, -1, -1, 69.5553097345
+7783320848.jpg, -1, -1, 72.373853211
+7785216770.jpg, -1, -1, 59.875
+7785223076.jpg, -1, -1, 14.2152777778
+7788745946.jpg, -1, -1, 72.8108695652
+7796133504.jpg, -1, -1, 71.3851351351
+7797268116.jpg, -1, -1, 64.8445945946
+7802695886.jpg, -1, -1, 75.1136363636
+7810074740.jpg, -1, -1, 75.3217391304
+7819211312.jpg, -1, -1, 50.1316964286
+7819666996.jpg, -1, -1, 72.722972973
+7832796652.jpg, -1, -1, 73.031512605
+7875370780.jpg, -1, -1, 67.640776699
+7877688978.jpg, -1, -1, 43.3782051282
+7880797072.jpg, -1, -1, 53.1985981308
+7889791504.jpg, -1, -1, 66.175
+7893560424.jpg, -1, -1, 57.8965517241
+7906332480.jpg, -1, -1, 65.2
+7911339166.jpg, -1, -1, 72.2629310345
+7911502338.jpg, -1, -1, 50.9380530973
+7919270808.jpg, -1, -1, 39.2828947368
+7927998516.jpg, -1, -1, 71.9054054054
+7928561010.jpg, -1, -1, 43.6116504854
+7934073070.jpg, -1, -1, 53.0984251969
+7937116986.jpg, -1, -1, 72.0711009174
+7942038522.jpg, -1, -1, 55.5089285714
+7946380974.jpg, -1, -1, 67.9930555556
+7954849148.jpg, -1, -1, 79.1822429907
+7955276034.jpg, -1, -1, 40.5558035714
+7955776040.jpg, -1, -1, 53.027027027
+7959317086.jpg, -1, -1, 60.2232142857
+7964545096.jpg, -1, -1, 20.5762711864
+7969586312.jpg, -1, -1, 66.342920354
+7973547250.jpg, -1, -1, 70.8782051282
+7974813040.jpg, -1, -1, 74.4893617021
+7978282630.jpg, -1, -1, 50.9852941176
+7983035556.jpg, -1, -1, 72.671875
+7991223182.jpg, -1, -1, 65.3648648649
+7991449337.jpg, -1, -1, 80.8509174312
+7994058325.jpg, -1, -1, 50.8855140187
+7999186610.jpg, -1, -1, 52.2522123894
+8000770201.jpg, -1, -1, 60.2428571429
+8003923043.jpg, -1, -1, 53.7847222222
+8005430981.jpg, -1, -1, 75.625
+8008341376.jpg, -1, -1, 66.7130434783
+8016728166.jpg, -1, -1, 70.3
+8017805417.jpg, -1, -1, 75.8970588235
+8020768393.jpg, -1, -1, 72.8191964286
+8021353022.jpg, -1, -1, 74.9553571429
+8021464988.jpg, -1, -1, 74.125
+8022493551.jpg, -1, -1, 57.7794117647
+8026326591.jpg, -1, -1, 72.2224770642
+8027425918.jpg, -1, -1, 75.1
+8030242803.jpg, -1, -1, 58.9086538462
+8034781018.jpg, -1, -1, 79.2614678899
+8035478082.jpg, -1, -1, 28.7570093458
+8036386438.jpg, -1, -1, 75.25
+8039894154.jpg, -1, -1, 41.2924107143
+8040007574.jpg, -1, -1, 78.9587155963
+80417582.jpg, -1, -1, 38.7836134454
+8044678667.jpg, -1, -1, 78.9166666667
+8056693769.jpg, -1, -1, 68.4732142857
+8064361291.jpg, -1, -1, 38.4806034483
+8065579848.jpg, -1, -1, 65.1747787611
+8070731836.jpg, -1, -1, 70.4314159292
+8078731890.jpg, -1, -1, 67.4419642857
+8081688318.jpg, -1, -1, 65.0287610619
+8087906672.jpg, -1, -1, 73.1138392857
+8088110789.jpg, -1, -1, 78.796728972
+8090361774.jpg, -1, -1, 35.9861111111
+8094310910.jpg, -1, -1, 72.8407079646
+8094870766.jpg, -1, -1, 71.3347457627
+8108912190.jpg, -1, -1, 67.2796610169
+8113190555.jpg, -1, -1, 78.2171052632
+8116678978.jpg, -1, -1, 78.5353982301
+8118290739.jpg, -1, -1, 57.6644736842
+8120735257.jpg, -1, -1, 72.7594339623
+8121424411.jpg, -1, -1, 43.9151376147
+8127311442.jpg, -1, -1, 47.1845794393
+81292981.jpg, -1, -1, 17.6375
+8130446271.jpg, -1, -1, 60.9547413793
+8130469512.jpg, -1, -1, 56.9719827586
+8142530403.jpg, -1, -1, 67.1434782609
+8142588958.jpg, -1, -1, 68.6941964286
+8152567908.jpg, -1, -1, 57.9180672269
+8154204617.jpg, -1, -1, 73.3461538462
+8157886874.jpg, -1, -1, 71.1961206897
+8158011851.jpg, -1, -1, 69.2239130435
+8158344434.jpg, -1, -1, 72.4276315789
+8163644568.jpg, -1, -1, 18.8172268908
+8178731250.jpg, -1, -1, 67.4159663866
+8179375247.jpg, -1, -1, 80.3455882353
+8179882827.jpg, -1, -1, 73.0347222222
+8183790186.jpg, -1, -1, 64.6959459459
+8184075727.jpg, -1, -1, 71.6725663717
+8185807831.jpg, -1, -1, 61.232300885
+8192678214.jpg, -1, -1, 75.6150442478
+8195543723.jpg, -1, -1, 76.7287735849
+8196266010.jpg, -1, -1, 36.5733944954
+8199545794.jpg, -1, -1, 77.6592920354
+8206370859.jpg, -1, -1, 77.9747706422
+8210857374.jpg, -1, -1, 67.3928571429
+8211686846.jpg, -1, -1, 52.3922018349
+8214901795.jpg, -1, -1, 67.3891509434
+8220742117.jpg, -1, -1, 20.125
+8226831484.jpg, -1, -1, 54.5538793103
+8229585759.jpg, -1, -1, 55.93125
+8230351643.jpg, -1, -1, 29.8341584158
+8231492407.jpg, -1, -1, 66.4107142857
+8238462203.jpg, -1, -1, 79.4096638655
+8239827069.jpg, -1, -1, 58.2
+8242811234.jpg, -1, -1, 66.3944954128
+8244017611.jpg, -1, -1, 62.2196261682
+8245350767.jpg, -1, -1, 74.2476635514
+8246184591.jpg, -1, -1, 50.15625
+8250498441.jpg, -1, -1, 60.8125
+8252742940.jpg, -1, -1, 83.5
+8256731117.jpg, -1, -1, 48.1118421053
+8257779790.jpg, -1, -1, 22.9533898305
+8259404043.jpg, -1, -1, 74.65
+8260840073.jpg, -1, -1, 49.5294117647
+8261560586.jpg, -1, -1, 65.4668141593
+826373.jpg, -1, -1, 77.7457983193
+8266041849.jpg, -1, -1, 68.1785714286
+8266648992.jpg, -1, -1, 73.3
+8267353806.jpg, -1, -1, 70.7566964286
+8269777387.jpg, -1, -1, 54.6642857143
+8281818195.jpg, -1, -1, 36.95625
+8282376526.jpg, -1, -1, 60.3394495413
+82824130.jpg, -1, -1, 29.4625
+8289927707.jpg, -1, -1, 66.1081081081
+8301271447.jpg, -1, -1, 63.5263157895
+8303486295.jpg, -1, -1, 77.2076271186
+8303662528.jpg, -1, -1, 65.6004464286
+8305212885.jpg, -1, -1, 44.5963302752
+8306155022.jpg, -1, -1, 77.65625
+8306770184.jpg, -1, -1, 65.3808411215
+8307817608.jpg, -1, -1, 33.5580357143
+8310254529.jpg, -1, -1, 71.19375
+8317893385.jpg, -1, -1, 75.18125
+8318268016.jpg, -1, -1, 65.4668141593
+8320017511.jpg, -1, -1, 77.0280172414
+8320098796.jpg, -1, -1, 55.2836134454
+8321155777.jpg, -1, -1, 53.2731092437
+8322413607.jpg, -1, -1, 72.9230769231
+8330807035.jpg, -1, -1, 34.9705882353
+8331758509.jpg, -1, -1, 68.3026315789
+8334203333.jpg, -1, -1, 48.0320512821
+8336846212.jpg, -1, -1, 68.375
+8337813363.jpg, -1, -1, 68.7331932773
+8338842182.jpg, -1, -1, 67.2920353982
+8346970641.jpg, -1, -1, 51.6891891892
+8349327834.jpg, -1, -1, 47.25
+8350532456.jpg, -1, -1, 66.8613445378
+8351745980.jpg, -1, -1, 61.5257009346
+8352587213.jpg, -1, -1, 63.8040540541
+8353048335.jpg, -1, -1, 62.2163865546
+8353276755.jpg, -1, -1, 71.125
+8354332729.jpg, -1, -1, 53.1052631579
+8354479425.jpg, -1, -1, 46.6351351351
+8357037213.jpg, -1, -1, 69.1949541284
+8358374077.jpg, -1, -1, 67.0
+8358381896.jpg, -1, -1, 57.85
+8362183655.jpg, -1, -1, 71.8714285714
+8365545870.jpg, -1, -1, 74.1680327869
+8374149596.jpg, -1, -1, 24.2079439252
+8377051997.jpg, -1, -1, 45.0491071429
+8380369304.jpg, -1, -1, 36.925
+8386378049.jpg, -1, -1, 66.6283783784
+8391500206.jpg, -1, -1, 46.2635135135
+8393106944.jpg, -1, -1, 22.2257281553
+8393175128.jpg, -1, -1, 68.9324324324
+8395711832.jpg, -1, -1, 38.3012820513
+8397860756.jpg, -1, -1, 56.3928571429
+8397958686.jpg, -1, -1, 60.6004672897
+8398198583.jpg, -1, -1, 78.231092437
+8399489944.jpg, -1, -1, 60.8653846154
+8401416435.jpg, -1, -1, 74.1903669725
+8402267016.jpg, -1, -1, 50.2027027027
+8403312891.jpg, -1, -1, 43.825
+8403342715.jpg, -1, -1, 48.8655660377
+8404206319.jpg, -1, -1, 62.4625
+8407650196.jpg, -1, -1, 60.7142857143
+8407751062.jpg, -1, -1, 57.1605504587
+8417650354.jpg, -1, -1, 69.3040540541
+8420816712.jpg, -1, -1, 42.3922413793
+8421980802.jpg, -1, -1, 66.8625
+8425720386.jpg, -1, -1, 54.7428571429
+8434015762.jpg, -1, -1, 49.15
+8434381156.jpg, -1, -1, 31.2689655172
+8434827647.jpg, -1, -1, 77.5657894737
+8435700290.jpg, -1, -1, 74.5870535714
+8435902776.jpg, -1, -1, 70.09375
+8437609150.jpg, -1, -1, 80.0379464286
+8438205347.jpg, -1, -1, 71.2028301887
+8438721601.jpg, -1, -1, 77.402173913
+8438943296.jpg, -1, -1, 53.6293103448
+8442013836.jpg, -1, -1, 46.784351145
+8442383096.jpg, -1, -1, 43.15
+8447352918.jpg, -1, -1, 43.6642857143
+8448710097.jpg, -1, -1, 60.6422018349
+8450680026.jpg, -1, -1, 61.9661016949
+8450836581.jpg, -1, -1, 70.5357142857
+8450870263.jpg, -1, -1, 56.2972972973
+8457028188.jpg, -1, -1, 77.858974359
+8458155212.jpg, -1, -1, 67.5593220339
+8460376605.jpg, -1, -1, 78.7857142857
+8461542458.jpg, -1, -1, 41.4097222222
+8462442319.jpg, -1, -1, 79.5608108108
+8463119506.jpg, -1, -1, 74.3411016949
+8463181410.jpg, -1, -1, 47.1116071429
+8465276999.jpg, -1, -1, 72.1918103448
+8465545790.jpg, -1, -1, 62.4509345794
+8473611231.jpg, -1, -1, 13.2593457944
+8474594407.jpg, -1, -1, 62.4080188679
+8481450412.jpg, -1, -1, 37.225
+8485295618.jpg, -1, -1, 50.95
+8486814802.jpg, -1, -1, 73.476635514
+8492768208.jpg, -1, -1, 71.3108108108
+8496923161.jpg, -1, -1, 65.9205607477
+8497829320.jpg, -1, -1, 62.0347222222
+8501012764.jpg, -1, -1, 51.2173913043
+8501698385.jpg, -1, -1, 69.0810810811
+8503871115.jpg, -1, -1, 60.4868421053
+8510431014.jpg, -1, -1, 58.9101941748
+85119046.jpg, -1, -1, 35.03125
+8515582085.jpg, -1, -1, 67.6875
+8518610606.jpg, -1, -1, 54.9196428571
+8522041587.jpg, -1, -1, 58.8973214286
+8524995451.jpg, -1, -1, 52.3581081081
+8527988312.jpg, -1, -1, 76.9717391304
+8528201518.jpg, -1, -1, 79.4864864865
+8528934790.jpg, -1, -1, 72.7894736842
+8529310804.jpg, -1, -1, 66.7847826087
+8529705971.jpg, -1, -1, 75.7610619469
+8530355188.jpg, -1, -1, 36.6432038835
+8531633123.jpg, -1, -1, 80.3325892857
+8532727626.jpg, -1, -1, 70.4663865546
+8533373653.jpg, -1, -1, 59.6184210526
+8535035859.jpg, -1, -1, 47.5586956522
+8535232822.jpg, -1, -1, 68.7522123894
+8539498079.jpg, -1, -1, 57.775
+8540532034.jpg, -1, -1, 52.6949541284
+8541006859.jpg, -1, -1, 61.328125
+8545722241.jpg, -1, -1, 52.5625
+8545905705.jpg, -1, -1, 29.875
+8550575599.jpg, -1, -1, 57.3621495327
+8552740610.jpg, -1, -1, 30.8211009174
+8555190491.jpg, -1, -1, 70.9391891892
+8556730555.jpg, -1, -1, 58.5186915888
+8556786880.jpg, -1, -1, 73.4247787611
+8558582535.jpg, -1, -1, 58.3340336134
+8560249138.jpg, -1, -1, 56.1631355932
+8565939221.jpg, -1, -1, 39.8108695652
+8567878134.jpg, -1, -1, 54.302173913
+8590374545.jpg, -1, -1, 66.1969026549
+8595948410.jpg, -1, -1, 33.4260869565
+8596901738.jpg, -1, -1, 34.3466386555
+8602839124.jpg, -1, -1, 44.1869565217
+8607324496.jpg, -1, -1, 50.3398058252
+8607422328.jpg, -1, -1, 24.44375
+8607616136.jpg, -1, -1, 43.2205882353
+8610015449.jpg, -1, -1, 69.775
+8610505282.jpg, -1, -1, 74.4180672269
+8618789190.jpg, -1, -1, 74.9444444444
+8623218161.jpg, -1, -1, 77.9760869565
+8626678643.jpg, -1, -1, 65.0086206897
+8630845484.jpg, -1, -1, 42.6118421053
+8631436677.jpg, -1, -1, 72.3428571429
+8632003704.jpg, -1, -1, 59.8451327434
+8634276040.jpg, -1, -1, 46.9097222222
+8634744119.jpg, -1, -1, 74.1805555556
+8639759293.jpg, -1, -1, 72.325
+8644207930.jpg, -1, -1, 26.4189189189
+8644571098.jpg, -1, -1, 76.7364864865
+8645014360.jpg, -1, -1, 44.0022123894
+8646019769.jpg, -1, -1, 66.3181818182
+8646711055.jpg, -1, -1, 72.5
+8648111033.jpg, -1, -1, 50.0805084746
+8655538784.jpg, -1, -1, 21.2162162162
+8664375707.jpg, -1, -1, 65.9205607477
+8667121921.jpg, -1, -1, 70.0394736842
+8668135949.jpg, -1, -1, 53.3986486486
+8668194198.jpg, -1, -1, 63.4324324324
+8670498239.jpg, -1, -1, 71.5412844037
+8675618794.jpg, -1, -1, 43.4391891892
+8676415201.jpg, -1, -1, 66.5580357143
+8683910174.jpg, -1, -1, 46.15
+8687755587.jpg, -1, -1, 57.7023809524
+8695646882.jpg, -1, -1, 77.24375
+8696567124.jpg, -1, -1, 35.0405405405
+8697364440.jpg, -1, -1, 66.2763157895
+8701800150.jpg, -1, -1, 71.3108108108
+8702450132.jpg, -1, -1, 57.5426829268
+8702518420.jpg, -1, -1, 76.9568965517
+8702530063.jpg, -1, -1, 68.8252212389
+8704133333.jpg, -1, -1, 70.5044247788
+8706343716.jpg, -1, -1, 60.5677966102
+8706810197.jpg, -1, -1, 70.6050420168
+8716882404.jpg, -1, -1, 69.2834821429
+8717130944.jpg, -1, -1, 72.7542016807
+8718842570.jpg, -1, -1, 56.5911214953
+8725613522.jpg, -1, -1, 33.1680672269
+8726161561.jpg, -1, -1, 74.3689320388
+8728752853.jpg, -1, -1, 67.222972973
+8729158424.jpg, -1, -1, 70.5357142857
+8730747598.jpg, -1, -1, 43.6880733945
+8731687035.jpg, -1, -1, 44.1081081081
+8732282616.jpg, -1, -1, 66.8413461538
+8742280365.jpg, -1, -1, 57.9385245902
+8742323176.jpg, -1, -1, 55.6995798319
+8743944148.jpg, -1, -1, 35.9324324324
+8746191422.jpg, -1, -1, 80.7759433962
+8746457937.jpg, -1, -1, 47.2876106195
+8751039261.jpg, -1, -1, 73.4474789916
+8751458979.jpg, -1, -1, 69.9203539823
+8754025405.jpg, -1, -1, 77.8326086957
+8754056878.jpg, -1, -1, 75.0290178571
+8755395714.jpg, -1, -1, 73.1695652174
+8755824682.jpg, -1, -1, 74.3247663551
+8772955624.jpg, -1, -1, 77.2586956522
+8776151757.jpg, -1, -1, 64.7702702703
+8797772647.jpg, -1, -1, 65.35
+8802725309.jpg, -1, -1, 40.724789916
+8816975901.jpg, -1, -1, 61.1071428571
+8817313814.jpg, -1, -1, 72.313559322
+8819567462.jpg, -1, -1, 40.1875
+8844582694.jpg, -1, -1, 49.035046729
+8862366940.jpg, -1, -1, 61.5491071429
+8864018641.jpg, -1, -1, 66.2887931034
+8882534571.jpg, -1, -1, 44.2775423729
+8899231880.jpg, -1, -1, 41.0389908257
+8901015763.jpg, -1, -1, 46.0803571429
+890231750.jpg, -1, -1, 49.9551886792
+8904864157.jpg, -1, -1, 68.5656934307
+8909361776.jpg, -1, -1, 68.4864864865
+8910371075.jpg, -1, -1, 74.7958715596
+8911730034.jpg, -1, -1, 60.8287401575
+8929900709.jpg, -1, -1, 62.3175675676
+8930797229.jpg, -1, -1, 81.1
+8933855427.jpg, -1, -1, 59.8935643564
+8938035476.jpg, -1, -1, 71.2385321101
+89393627.jpg, -1, -1, 27.4138655462
+8965909066.jpg, -1, -1, 68.7331932773
+8975620960.jpg, -1, -1, 50.8683035714
+8984346211.jpg, -1, -1, 69.0442477876
+8984699375.jpg, -1, -1, 55.140625
+9019428727.jpg, -1, -1, 66.925
+9029384629.jpg, -1, -1, 30.2641509434
+9034047597.jpg, -1, -1, 65.7909482759
+9034707954.jpg, -1, -1, 57.2302631579
+9034716133.jpg, -1, -1, 77.2546728972
+9036180286.jpg, -1, -1, 72.3035714286
+9050865219.jpg, -1, -1, 75.0972222222
+9052518863.jpg, -1, -1, 52.975
+9055879747.jpg, -1, -1, 54.2844036697
+9058722713.jpg, -1, -1, 55.3940677966
+9069424514.jpg, -1, -1, 56.8224299065
+9070049178.jpg, -1, -1, 73.825
+9076304306.jpg, -1, -1, 34.1386554622
+9102604497.jpg, -1, -1, 41.8608490566
+9109271301.jpg, -1, -1, 71.7803738318
+9109755543.jpg, -1, -1, 28.3151260504
+9113486865.jpg, -1, -1, 76.6378504673
+9115676276.jpg, -1, -1, 70.6330275229
+9119961649.jpg, -1, -1, 81.1025641026
+9121896285.jpg, -1, -1, 57.6351351351
+9131687822.jpg, -1, -1, 73.9586956522
+91357025.jpg, -1, -1, 64.5042016807
+9141403760.jpg, -1, -1, 62.125
+9150255535.jpg, -1, -1, 78.4583333333
+9159337218.jpg, -1, -1, 74.7
+9186366283.jpg, -1, -1, 61.0
+9194756449.jpg, -1, -1, 48.1927966102
+9198766588.jpg, -1, -1, 72.7894736842
+9198779264.jpg, -1, -1, 45.0491071429
+9200829224.jpg, -1, -1, 73.8628318584
+9202257544.jpg, -1, -1, 63.2569444444
+9209619689.jpg, -1, -1, 68.2090517241
+9212037816.jpg, -1, -1, 71.2013888889
+9220351854.jpg, -1, -1, 60.6911764706
+9221345159.jpg, -1, -1, 62.7016806723
+9225734576.jpg, -1, -1, 67.5110619469
+9225878690.jpg, -1, -1, 55.0517241379
+9226793819.jpg, -1, -1, 32.5441176471
+9232503869.jpg, -1, -1, 32.8100961538
+9234807959.jpg, -1, -1, 47.0150862069
+9238616212.jpg, -1, -1, 70.2004310345
+9239862860.jpg, -1, -1, 66.9276315789
+9245144961.jpg, -1, -1, 59.7721238938
+92466227.jpg, -1, -1, 48.2121848739
+9249859612.jpg, -1, -1, 67.9082568807
+9269195837.jpg, -1, -1, 64.5472972973
+9274677525.jpg, -1, -1, 58.0567226891
+9277184417.jpg, -1, -1, 53.8883928571
+9295408685.jpg, -1, -1, 58.5091240876
+9299238943.jpg, -1, -1, 49.737394958
+930643510.jpg, -1, -1, 24.1148648649
+9310356178.jpg, -1, -1, 58.3040540541
+9315620500.jpg, -1, -1, 69.0342465753
+9322027317.jpg, -1, -1, 58.75
+9324654543.jpg, -1, -1, 61.456
+9329940580.jpg, -1, -1, 50.95
+9333668178.jpg, -1, -1, 74.1351351351
+9334502001.jpg, -1, -1, 53.0328947368
+9341616816.jpg, -1, -1, 75.25
+9346372358.jpg, -1, -1, 74.8080357143
+9362014571.jpg, -1, -1, 45.4782608696
+9365989820.jpg, -1, -1, 55.2476415094
+9369620513.jpg, -1, -1, 76.8705357143
+9372534946.jpg, -1, -1, 61.7972972973
+9375359458.jpg, -1, -1, 65.8214285714
+9379361831.jpg, -1, -1, 77.3861607143
+9401118786.jpg, -1, -1, 52.5986842105
+940371579.jpg, -1, -1, 59.50625
+9408400121.jpg, -1, -1, 71.5756302521
+9419183210.jpg, -1, -1, 69.936440678
+9435287371.jpg, -1, -1, 78.9407894737
+9439902792.jpg, -1, -1, 75.9571428571
+9439964500.jpg, -1, -1, 55.4811320755
+944341845.jpg, -1, -1, 52.1952054795
+9463363154.jpg, -1, -1, 63.9264705882
+9468744508.jpg, -1, -1, 79.375
+9472868070.jpg, -1, -1, 61.4471153846
+9476849940.jpg, -1, -1, 73.4705882353
+9480702491.jpg, -1, -1, 73.3622881356
+9482519952.jpg, -1, -1, 57.5047169811
+9483838441.jpg, -1, -1, 78.4143835616
+9488096665.jpg, -1, -1, 76.6012931034
+9488864072.jpg, -1, -1, 50.6586538462
+9489836089.jpg, -1, -1, 68.4380733945
+9493293324.jpg, -1, -1, 72.3552631579
+9500372295.jpg, -1, -1, 44.3125
+9505522131.jpg, -1, -1, 66.5769230769
+9510873691.jpg, -1, -1, 67.5110619469
+952167260.jpg, -1, -1, 48.3846153846
+9523882653.jpg, -1, -1, 47.5928571429
+9524699151.jpg, -1, -1, 77.3304347826
+9527678556.jpg, -1, -1, 65.7428571429
+9527980922.jpg, -1, -1, 70.7430555556
+9536683213.jpg, -1, -1, 65.5212264151
+9537434864.jpg, -1, -1, 59.1150442478
+9538674147.jpg, -1, -1, 45.6559633028
+9541231306.jpg, -1, -1, 24.7368421053
+9541618071.jpg, -1, -1, 41.8157894737
+9581336227.jpg, -1, -1, 58.1607142857
+9582086369.jpg, -1, -1, 75.6086956522
+9585740831.jpg, -1, -1, 70.5560344828
+9586652263.jpg, -1, -1, 74.9527027027
+9587630414.jpg, -1, -1, 75.31875
+9590556429.jpg, -1, -1, 34.4159663866
+9602714230.jpg, -1, -1, 60.6081081081
+9604185976.jpg, -1, -1, 75.25
+9605303857.jpg, -1, -1, 67.0
+9613807416.jpg, -1, -1, 79.5223214286
+9630325643.jpg, -1, -1, 32.2591743119
+9639418449.jpg, -1, -1, 69.2956521739
+9646776074.jpg, -1, -1, 67.5110619469
+9650999310.jpg, -1, -1, 43.6365546218
+9674243129.jpg, -1, -1, 63.2155963303
+9681377633.jpg, -1, -1, 18.0298165138
+9684211981.jpg, -1, -1, 75.7352941176
+9686480737.jpg, -1, -1, 26.0019083969
+9687081006.jpg, -1, -1, 66.4701834862
+9694884270.jpg, -1, -1, 54.1011904762
+9702019166.jpg, -1, -1, 60.0135135135
+970460687.jpg, -1, -1, 65.6716101695
+9710823281.jpg, -1, -1, 65.2714285714
+9712051180.jpg, -1, -1, 69.4357142857
+9712191760.jpg, -1, -1, 64.8097345133
+9714595106.jpg, -1, -1, 48.1637168142
+9722821291.jpg, -1, -1, 32.7415254237
+9723956053.jpg, -1, -1, 64.9
+9724373732.jpg, -1, -1, 65.8541666667
+9731493509.jpg, -1, -1, 27.6597222222
+9733980007.jpg, -1, -1, 43.6495535714
+9737673243.jpg, -1, -1, 58.0291262136
+9759154326.jpg, -1, -1, 60.0783898305
+9762410596.jpg, -1, -1, 70.9152542373
+9765112254.jpg, -1, -1, 59.0296610169
+9766865862.jpg, -1, -1, 48.0872641509
+9770742411.jpg, -1, -1, 70.6
+9774901765.jpg, -1, -1, 46.6399082569
+9780398791.jpg, -1, -1, 68.4102564103
+9784668705.jpg, -1, -1, 70.9152542373
+9788787724.jpg, -1, -1, 28.2027027027
+97984449.jpg, -1, -1, 47.0336134454
+9829643193.jpg, -1, -1, 69.1554054054
+9834223025.jpg, -1, -1, 57.0405405405
+9859562873.jpg, -1, -1, 36.2834821429
+9866692876.jpg, -1, -1, 68.1092436975
+9877078626.jpg, -1, -1, 75.8846153846
+9884648494.jpg, -1, -1, 71.8940677966
+9895608016.jpg, -1, -1, 52.0197368421
+9906667675.jpg, -1, -1, 35.468220339
+9910043795.jpg, -1, -1, 72.8407079646
+9910216783.jpg, -1, -1, 78.4623893805
+9913105155.jpg, -1, -1, 72.452173913
+9915601385.jpg, -1, -1, 65.6602564103
+9916890605.jpg, -1, -1, 56.1285046729
+9923882853.jpg, -1, -1, 54.6949152542
+9925803176.jpg, -1, -1, 46.5197368421
+9935149655.jpg, -1, -1, 53.6730769231
+9936084895.jpg, -1, -1, 61.6486486486
+9977056676.jpg, -1, -1, 46.3013392857
+9977191265.jpg, -1, -1, 74.4716981132
+9984535544.jpg, -1, -1, 70.0200892857
+9991999836.jpg, -1, -1, 78.0924369748

examplar_data_labels/KoNiQ10k/validation_labels.txt

@@ -0,0 +1,1000 @@
+10043785683.jpg, -1, -1, 71.2043269231
+10059344614.jpg, -1, -1, 68.6651376147
+10061596254.jpg, -1, -1, 59.6418918919
+10067457634.jpg, -1, -1, 78.4390756303
+10082923485.jpg, -1, -1, 53.2260869565
+10089177336.jpg, -1, -1, 72.9459459459
+10140071304.jpg, -1, -1, 62.5576923077
+10164842933.jpg, -1, -1, 76.69375
+10241895785.jpg, -1, -1, 67.0736607143
+10244314384.jpg, -1, -1, 47.5481651376
+10246311914.jpg, -1, -1, 75.4579831933
+10264804094.jpg, -1, -1, 76.6495535714
+10264984793.jpg, -1, -1, 66.1
+10344523015.jpg, -1, -1, 52.5263157895
+10371125786.jpg, -1, -1, 38.2451456311
+10409059825.jpg, -1, -1, 73.1513157895
+10414716033.jpg, -1, -1, 71.0878378378
+10415631353.jpg, -1, -1, 43.0819327731
+10453950204.jpg, -1, -1, 58.975
+10458591.jpg, -1, -1, 51.4871794872
+10463545414.jpg, -1, -1, 55.0444915254
+10465121613.jpg, -1, -1, 61.5723684211
+10466350116.jpg, -1, -1, 77.2243589744
+10492897986.jpg, -1, -1, 44.6283783784
+10513907744.jpg, -1, -1, 59.4311926606
+10544833625.jpg, -1, -1, 50.5756880734
+10567188645.jpg, -1, -1, 64.1575630252
+10605846204.jpg, -1, -1, 70.4620535714
+10619909515.jpg, -1, -1, 60.7775423729
+10628497685.jpg, -1, -1, 54.947826087
+10680502673.jpg, -1, -1, 78.75625
+10686332213.jpg, -1, -1, 57.4216101695
+10689538413.jpg, -1, -1, 73.66875
+10692654005.jpg, -1, -1, 68.2
+10692667386.jpg, -1, -1, 71.9033018868
+10694350076.jpg, -1, -1, 73.1285714286
+10706770686.jpg, -1, -1, 76.8519417476
+10728525673.jpg, -1, -1, 68.3026315789
+107505789.jpg, -1, -1, 54.35
+10770026303.jpg, -1, -1, 72.0825892857
+10784159775.jpg, -1, -1, 74.8849557522
+1085590803.jpg, -1, -1, 24.0558035714
+10856757.jpg, -1, -1, 32.2954545455
+10883456586.jpg, -1, -1, 28.8793103448
+10905114716.jpg, -1, -1, 50.0695652174
+10923819485.jpg, -1, -1, 40.8275862069
+10947519133.jpg, -1, -1, 74.609223301
+10990887046.jpg, -1, -1, 58.0765306122
+11062902804.jpg, -1, -1, 69.9117647059
+11068654734.jpg, -1, -1, 68.0694444444
+11148529964.jpg, -1, -1, 58.2
+11170786154.jpg, -1, -1, 67.6932773109
+11172371243.jpg, -1, -1, 75.8785714286
+11229777185.jpg, -1, -1, 66.0747663551
+11256485296.jpg, -1, -1, 61.9044117647
+11262156493.jpg, -1, -1, 79.9271653543
+11268202995.jpg, -1, -1, 41.734375
+11273127766.jpg, -1, -1, 79.5575221239
+11291450085.jpg, -1, -1, 56.7394957983
+1130601148.jpg, -1, -1, 72.3804347826
+11333447674.jpg, -1, -1, 76.9978813559
+11333782783.jpg, -1, -1, 70.3584070796
+11338371864.jpg, -1, -1, 68.8245192308
+11341117543.jpg, -1, -1, 61.6850961538
+11369797883.jpg, -1, -1, 74.0277777778
+11372421485.jpg, -1, -1, 65.685840708
+11396447303.jpg, -1, -1, 48.0107758621
+11414474354.jpg, -1, -1, 18.5514705882
+11426976205.jpg, -1, -1, 35.188559322
+11433593823.jpg, -1, -1, 62.9486607143
+11436378704.jpg, -1, -1, 79.3397435897
+11474376103.jpg, -1, -1, 53.25
+11484412296.jpg, -1, -1, 63.2641509434
+11518631385.jpg, -1, -1, 32.2668067227
+11533404563.jpg, -1, -1, 65.4668141593
+11541375264.jpg, -1, -1, 50.3513513514
+11555549865.jpg, -1, -1, 60.9402654867
+117172917.jpg, -1, -1, 62.9789915966
+122553631.jpg, -1, -1, 37.85
+128781220.jpg, -1, -1, 59.7205882353
+1297819816.jpg, -1, -1, 74.2794117647
+1316239068.jpg, -1, -1, 61.6165254237
+1396109073.jpg, -1, -1, 63.4343220339
+1402675396.jpg, -1, -1, 29.9102564103
+142117255.jpg, -1, -1, 67.9625
+1434258458.jpg, -1, -1, 67.8608695652
+145215893.jpg, -1, -1, 10.14375
+147639760.jpg, -1, -1, 47.40625
+149836125.jpg, -1, -1, 46.756302521
+1540229828.jpg, -1, -1, 72.9125
+1543929747.jpg, -1, -1, 33.7203389831
+1561250178.jpg, -1, -1, 61.4191176471
+160999577.jpg, -1, -1, 30.5909090909
+1626230781.jpg, -1, -1, 15.2293577982
+1799451312.jpg, -1, -1, 62.0945945946
+183156510.jpg, -1, -1, 48.8361344538
+186520031.jpg, -1, -1, 60.8298319328
+1947195937.jpg, -1, -1, 41.0818584071
+197324620.jpg, -1, -1, 62.5630252101
+1995190133.jpg, -1, -1, 77.369266055
+208949930.jpg, -1, -1, 72.6848739496
+2115915075.jpg, -1, -1, 55.8086956522
+2149719899.jpg, -1, -1, 38.1621621622
+2172409998.jpg, -1, -1, 56.0
+2190310.jpg, -1, -1, 58.3269230769
+2197636094.jpg, -1, -1, 72.286407767
+2199106763.jpg, -1, -1, 50.775
+2203797715.jpg, -1, -1, 76.3096330275
+2204368170.jpg, -1, -1, 44.2775423729
+2217543976.jpg, -1, -1, 76.1339285714
+2221745311.jpg, -1, -1, 39.110619469
+2223255997.jpg, -1, -1, 46.825
+2225683996.jpg, -1, -1, 73.5168067227
+2227011292.jpg, -1, -1, 70.6580188679
+2237630871.jpg, -1, -1, 78.55
+2239403255.jpg, -1, -1, 77.4711538462
+2255689742.jpg, -1, -1, 58.75
+2260527943.jpg, -1, -1, 58.4472477064
+2263434837.jpg, -1, -1, 77.525862069
+2280432289.jpg, -1, -1, 43.4836448598
+2293775114.jpg, -1, -1, 43.6483050847
+22947511.jpg, -1, -1, 13.9743589744
+2305236722.jpg, -1, -1, 25.4605263158
+2306462251.jpg, -1, -1, 35.0405405405
+2308986836.jpg, -1, -1, 64.6869158879
+2314346258.jpg, -1, -1, 54.2754237288
+2322053874.jpg, -1, -1, 80.0336134454
+232327907.jpg, -1, -1, 57.0861344538
+2325353113.jpg, -1, -1, 14.5882352941
+2344211137.jpg, -1, -1, 68.6351351351
+2355891811.jpg, -1, -1, 34.9082568807
+2363717244.jpg, -1, -1, 75.5296610169
+2367261033.jpg, -1, -1, 18.3175675676
+2367765365.jpg, -1, -1, 44.3486842105
+2405333928.jpg, -1, -1, 51.0593220339
+2415583702.jpg, -1, -1, 29.3156779661
+2416231204.jpg, -1, -1, 57.5921052632
+2433641656.jpg, -1, -1, 55.2455752212
+2437860365.jpg, -1, -1, 71.725
+2447797924.jpg, -1, -1, 41.65
+2460847286.jpg, -1, -1, 36.3349056604
+2476813465.jpg, -1, -1, 54.5114678899
+2500618710.jpg, -1, -1, 71.0891304348
+2503869220.jpg, -1, -1, 42.0186915888
+2519479147.jpg, -1, -1, 66.5657894737
+2538038277.jpg, -1, -1, 68.8025210084
+2542492471.jpg, -1, -1, 40.8869565217
+255857546.jpg, -1, -1, 75.1806722689
+2569803000.jpg, -1, -1, 57.8815789474
+2636589178.jpg, -1, -1, 54.6973684211
+2654797164.jpg, -1, -1, 71.6357142857
+2664892520.jpg, -1, -1, 54.5939849624
+26847054.jpg, -1, -1, 58.6794871795
+2688019250.jpg, -1, -1, 52.8571428571
+2703740653.jpg, -1, -1, 61.2642857143
+2726774995.jpg, -1, -1, 27.55
+2741661469.jpg, -1, -1, 69.0046728972
+274223521.jpg, -1, -1, 32.9976635514
+2743703757.jpg, -1, -1, 52.75
+2748301393.jpg, -1, -1, 40.825
+278636331.jpg, -1, -1, 58.0870535714
+2807676440.jpg, -1, -1, 63.175
+2924727768.jpg, -1, -1, 67.2920353982
+2938430307.jpg, -1, -1, 72.5982142857
+2941933060.jpg, -1, -1, 79.4789915966
+2946661592.jpg, -1, -1, 71.0210084034
+2952066894.jpg, -1, -1, 70.2702702703
+2986131608.jpg, -1, -1, 50.5723684211
+30054528.jpg, -1, -1, 69.3072033898
+3006180799.jpg, -1, -1, 55.1824324324
+3007216990.jpg, -1, -1, 39.1302521008
+3048643650.jpg, -1, -1, 53.2283464567
+3049913075.jpg, -1, -1, 71.2103448276
+3069379747.jpg, -1, -1, 63.1553398058
+3091624267.jpg, -1, -1, 4.94871794872
+3092999912.jpg, -1, -1, 33.55
+3115828038.jpg, -1, -1, 24.4928571429
+3125804438.jpg, -1, -1, 56.4125
+3128928410.jpg, -1, -1, 72.5252293578
+3130276162.jpg, -1, -1, 41.63125
+3132104449.jpg, -1, -1, 74.8571428571
+3146365512.jpg, -1, -1, 79.2654867257
+3170684406.jpg, -1, -1, 69.8424369748
+3193120688.jpg, -1, -1, 61.6334951456
+3195516092.jpg, -1, -1, 69.4092920354
+3201560837.jpg, -1, -1, 41.734375
+3228957282.jpg, -1, -1, 69.5282258065
+324220404.jpg, -1, -1, 62.6100917431
+3259283855.jpg, -1, -1, 45.4623893805
+3281066235.jpg, -1, -1, 73.0911214953
+3284203868.jpg, -1, -1, 70.0
+3304676774.jpg, -1, -1, 63.4642857143
+3323937813.jpg, -1, -1, 58.75
+3347228837.jpg, -1, -1, 70.7714285714
+3351610329.jpg, -1, -1, 65.8950892857
+3355714233.jpg, -1, -1, 58.8898305085
+336052627.jpg, -1, -1, 42.6216216216
+3364868711.jpg, -1, -1, 26.5675675676
+3367803108.jpg, -1, -1, 66.9221698113
+3379476157.jpg, -1, -1, 62.4555084746
+338924292.jpg, -1, -1, 32.317961165
+3405525198.jpg, -1, -1, 34.2115384615
+3411974393.jpg, -1, -1, 50.5
+3412359483.jpg, -1, -1, 67.9545454545
+3420154563.jpg, -1, -1, 52.7372881356
+3421464878.jpg, -1, -1, 63.5940366972
+3437594883.jpg, -1, -1, 34.7690677966
+3450351361.jpg, -1, -1, 82.3318584071
+3468424403.jpg, -1, -1, 54.3453389831
+3485554199.jpg, -1, -1, 50.3457943925
+3490456545.jpg, -1, -1, 38.9639830508
+3494678423.jpg, -1, -1, 48.2214285714
+3502153355.jpg, -1, -1, 63.860619469
+35311267.jpg, -1, -1, 26.9978991597
+3532282772.jpg, -1, -1, 60.3857758621
+3547062481.jpg, -1, -1, 64.0535714286
+3549424661.jpg, -1, -1, 41.0125
+3566079728.jpg, -1, -1, 63.1351351351
+3575258454.jpg, -1, -1, 64.0142857143
+3576771116.jpg, -1, -1, 52.997706422
+3577464388.jpg, -1, -1, 67.6629464286
+3601352242.jpg, -1, -1, 66.9294871795
+3603046487.jpg, -1, -1, 49.3717948718
+3604508101.jpg, -1, -1, 63.55
+3620678223.jpg, -1, -1, 59.2041284404
+364777572.jpg, -1, -1, 63.2433035714
+3658535721.jpg, -1, -1, 78.0
+3658667754.jpg, -1, -1, 69.8448275862
+3662491369.jpg, -1, -1, 48.2121848739
+3662558151.jpg, -1, -1, 45.8073394495
+3663369776.jpg, -1, -1, 74.6260504202
+3670445078.jpg, -1, -1, 31.7397260274
+3689872234.jpg, -1, -1, 40.875
+3697493807.jpg, -1, -1, 82.2909482759
+3697530291.jpg, -1, -1, 82.0064655172
+3707022590.jpg, -1, -1, 81.5256410256
+3715780832.jpg, -1, -1, 61.8347826087
+3736336896.jpg, -1, -1, 49.8310810811
+3741259519.jpg, -1, -1, 66.342920354
+374612387.jpg, -1, -1, 64.1334745763
+375446021.jpg, -1, -1, 48.2165178571
+3775591333.jpg, -1, -1, 43.6378504673
+3805107589.jpg, -1, -1, 59.4612068966
+3824105113.jpg, -1, -1, 55.8834745763
+3824485636.jpg, -1, -1, 48.1625
+3827911032.jpg, -1, -1, 72.3971962617
+3843859642.jpg, -1, -1, 65.8214285714
+3846071699.jpg, -1, -1, 66.3503937008
+3847022878.jpg, -1, -1, 66.475
+3876523676.jpg, -1, -1, 72.2105263158
+3891717415.jpg, -1, -1, 71.4196428571
+3895663827.jpg, -1, -1, 46.6805555556
+3904942577.jpg, -1, -1, 18.2173913043
+3918062562.jpg, -1, -1, 76.7435344828
+392179845.jpg, -1, -1, 16.2482758621
+3932630945.jpg, -1, -1, 61.6227678571
+3946405001.jpg, -1, -1, 71.6228448276
+3948277105.jpg, -1, -1, 57.9789719626
+3952840501.jpg, -1, -1, 76.0
+3955560376.jpg, -1, -1, 63.0847457627
+3960128594.jpg, -1, -1, 53.5555555556
+3966829877.jpg, -1, -1, 68.0760869565
+3976883946.jpg, -1, -1, 56.5265957447
+3980401611.jpg, -1, -1, 45.4782608696
+4000012271.jpg, -1, -1, 77.45
+4005014529.jpg, -1, -1, 56.1067961165
+4008219206.jpg, -1, -1, 57.6041666667
+4016398417.jpg, -1, -1, 61.2544642857
+4016639569.jpg, -1, -1, 66.0508849558
+4023245742.jpg, -1, -1, 53.5504201681
+4023598688.jpg, -1, -1, 53.0733944954
+4029731030.jpg, -1, -1, 50.95
+4037555676.jpg, -1, -1, 24.1419491525
+404033442.jpg, -1, -1, 62.8050847458
+4046716515.jpg, -1, -1, 41.0534351145
+405908679.jpg, -1, -1, 65.2281879195
+4067961238.jpg, -1, -1, 40.0879310345
+4068780416.jpg, -1, -1, 43.9868421053
+4078925567.jpg, -1, -1, 44.6539735099
+4086204414.jpg, -1, -1, 27.625
+409233389.jpg, -1, -1, 46.0576923077
+4095383312.jpg, -1, -1, 39.4495412844
+4099198318.jpg, -1, -1, 49.7431192661
+4100631766.jpg, -1, -1, 42.0986238532
+4106517753.jpg, -1, -1, 75.108974359
+4110400471.jpg, -1, -1, 59.7543478261
+411787882.jpg, -1, -1, 75.4652173913
+4127387822.jpg, -1, -1, 75.5344827586
+4132558989.jpg, -1, -1, 75.8340707965
+4138608778.jpg, -1, -1, 65.7415254237
+415283727.jpg, -1, -1, 49.9889380531
+4162338388.jpg, -1, -1, 66.5373831776
+418145002.jpg, -1, -1, 58.6714285714
+4193122125.jpg, -1, -1, 59.725
+4193696727.jpg, -1, -1, 56.1631355932
+4194035905.jpg, -1, -1, 70.6418918919
+4214550696.jpg, -1, -1, 56.8130434783
+4217243298.jpg, -1, -1, 55.7566371681
+4237536969.jpg, -1, -1, 75.4206896552
+4259140344.jpg, -1, -1, 54.6635514019
+4270574025.jpg, -1, -1, 39.175
+4299890430.jpg, -1, -1, 62.6148648649
+430051470.jpg, -1, -1, 62.2857142857
+4304984630.jpg, -1, -1, 63.7995689655
+4305975001.jpg, -1, -1, 73.9977678571
+4311062470.jpg, -1, -1, 64.5394736842
+4320399633.jpg, -1, -1, 61.7075471698
+4336460350.jpg, -1, -1, 75.3236607143
+4338156096.jpg, -1, -1, 64.421875
+434605634.jpg, -1, -1, 70.0581896552
+4347675881.jpg, -1, -1, 50.0769230769
+4365387751.jpg, -1, -1, 67.2946428571
+4370988573.jpg, -1, -1, 39.0609243697
+4388983100.jpg, -1, -1, 32.5878378378
+4408351965.jpg, -1, -1, 61.1681034483
+441241455.jpg, -1, -1, 45.0
+4423525827.jpg, -1, -1, 47.7242990654
+4430414110.jpg, -1, -1, 79.3403361345
+4432502474.jpg, -1, -1, 59.5068807339
+4435718106.jpg, -1, -1, 55.3529411765
+4436002363.jpg, -1, -1, 51.2208737864
+4448002516.jpg, -1, -1, 71.2364864865
+4463467980.jpg, -1, -1, 80.81875
+4471526544.jpg, -1, -1, 61.8262711864
+4480082097.jpg, -1, -1, 76.3782051282
+450063799.jpg, -1, -1, 61.3472222222
+450338979.jpg, -1, -1, 7.27876106195
+4512123861.jpg, -1, -1, 74.2423664122
+4512474932.jpg, -1, -1, 74.8680555556
+4525459727.jpg, -1, -1, 76.6
+4530384020.jpg, -1, -1, 63.6214285714
+4545054058.jpg, -1, -1, 24.7012711864
+4550104555.jpg, -1, -1, 69.4264705882
+4562682674.jpg, -1, -1, 39.8928571429
+4565683087.jpg, -1, -1, 80.3680555556
+4571847379.jpg, -1, -1, 53.1517857143
+4571937568.jpg, -1, -1, 56.2894736842
+4585233162.jpg, -1, -1, 43.9717391304
+4586793310.jpg, -1, -1, 50.5693277311
+4602769821.jpg, -1, -1, 75.3886554622
+4617163865.jpg, -1, -1, 28.9044117647
+4631640521.jpg, -1, -1, 73.7168141593
+463426919.jpg, -1, -1, 35.2478991597
+4638196012.jpg, -1, -1, 73.7898230088
+4669754441.jpg, -1, -1, 73.6717391304
+4683892759.jpg, -1, -1, 52.4275700935
+4688672987.jpg, -1, -1, 61.2943925234
+4691973869.jpg, -1, -1, 34.0
+4698449914.jpg, -1, -1, 43.7891791045
+4698931028.jpg, -1, -1, 65.5192307692
+4706726398.jpg, -1, -1, 40.935840708
+47095522.jpg, -1, -1, 67.4194915254
+4710464067.jpg, -1, -1, 73.3918918919
+4714778162.jpg, -1, -1, 65.6642857143
+4714961173.jpg, -1, -1, 66.2860576923
+4718032191.jpg, -1, -1, 58.8205128205
+4728497082.jpg, -1, -1, 70.8453389831
+4733042196.jpg, -1, -1, 73.6282051282
+4735563555.jpg, -1, -1, 68.3623853211
+4750416133.jpg, -1, -1, 65.6016949153
+4761802851.jpg, -1, -1, 50.1972477064
+4781552174.jpg, -1, -1, 68.2110091743
+4782546865.jpg, -1, -1, 67.0693277311
+4782987041.jpg, -1, -1, 79.03125
+4786690796.jpg, -1, -1, 36.5651260504
+4787504766.jpg, -1, -1, 62.1513157895
+4791175187.jpg, -1, -1, 61.1764705882
+4804146129.jpg, -1, -1, 39.8819444444
+4819833644.jpg, -1, -1, 23.6008403361
+4840518597.jpg, -1, -1, 56.1766055046
+4846003639.jpg, -1, -1, 57.9385245902
+4847954914.jpg, -1, -1, 54.49609375
+486170909.jpg, -1, -1, 56.5401785714
+4877994171.jpg, -1, -1, 54.2905405405
+4883254364.jpg, -1, -1, 68.3258928571
+4886277173.jpg, -1, -1, 30.025
+4886518096.jpg, -1, -1, 50.9230769231
+4887754576.jpg, -1, -1, 51.0110619469
+4890086996.jpg, -1, -1, 59.9911504425
+4900565313.jpg, -1, -1, 49.4485294118
+4908270930.jpg, -1, -1, 46.3371559633
+4912317009.jpg, -1, -1, 76.2617924528
+4918979865.jpg, -1, -1, 68.0760869565
+4921474940.jpg, -1, -1, 63.0306603774
+4922576334.jpg, -1, -1, 68.4601769912
+4924398774.jpg, -1, -1, 51.4575892857
+4926447658.jpg, -1, -1, 59.9610091743
+4936086328.jpg, -1, -1, 73.1875
+494015270.jpg, -1, -1, 50.9541284404
+4950276392.jpg, -1, -1, 39.4768907563
+4952399474.jpg, -1, -1, 59.8260869565
+4953755434.jpg, -1, -1, 58.45
+4975468754.jpg, -1, -1, 63.745412844
+4976142964.jpg, -1, -1, 52.6554054054
+4980059014.jpg, -1, -1, 60.2867647059
+498265046.jpg, -1, -1, 36.8875
+499095228.jpg, -1, -1, 81.9862385321
+4992884384.jpg, -1, -1, 70.3846153846
+4996581735.jpg, -1, -1, 78.6470588235
+5001719293.jpg, -1, -1, 36.936440678
+5003482891.jpg, -1, -1, 67.4194915254
+5005406987.jpg, -1, -1, 67.6991525424
+5008940426.jpg, -1, -1, 59.8782051282
+5012894248.jpg, -1, -1, 65.9239130435
+5016823142.jpg, -1, -1, 49.2184466019
+5024762120.jpg, -1, -1, 76.2162162162
+5026427468.jpg, -1, -1, 64.0586956522
+5032946224.jpg, -1, -1, 40.9915254237
+5040853192.jpg, -1, -1, 23.7896039604
+5043851130.jpg, -1, -1, 38.5216346154
+5062026548.jpg, -1, -1, 38.5592105263
+5065354809.jpg, -1, -1, 67.7795275591
+5078922181.jpg, -1, -1, 56.171875
+5081637571.jpg, -1, -1, 50.9852941176
+5083763203.jpg, -1, -1, 42.7610619469
+5091924028.jpg, -1, -1, 25.8308823529
+5092945872.jpg, -1, -1, 28.775
+5097621606.jpg, -1, -1, 62.2316513761
+5105331099.jpg, -1, -1, 15.8058035714
+5107763844.jpg, -1, -1, 65.3823529412
+5117589590.jpg, -1, -1, 69.5108695652
+5118376207.jpg, -1, -1, 65.05
+5124446081.jpg, -1, -1, 64.0882352941
+5130720224.jpg, -1, -1, 73.6458333333
+5132091056.jpg, -1, -1, 41.0088495575
+513723164.jpg, -1, -1, 56.1216814159
+5165838001.jpg, -1, -1, 71.7243589744
+5170858607.jpg, -1, -1, 65.3794642857
+517369320.jpg, -1, -1, 60.0
+5175429795.jpg, -1, -1, 74.7958715596
+5184701000.jpg, -1, -1, 24.5304347826
+5197308883.jpg, -1, -1, 69.3072033898
+5211888253.jpg, -1, -1, 28.7775229358
+5212639990.jpg, -1, -1, 60.4908256881
+5224316825.jpg, -1, -1, 30.9158878505
+5235352004.jpg, -1, -1, 42.6880530973
+5240740919.jpg, -1, -1, 61.2544642857
+5243003019.jpg, -1, -1, 43.8580508475
+52568944.jpg, -1, -1, 29.0215517241
+5261084727.jpg, -1, -1, 68.2411504425
+5266123407.jpg, -1, -1, 74.0409482759
+5270171351.jpg, -1, -1, 41.1041666667
+5275524338.jpg, -1, -1, 74.8680555556
+5286915372.jpg, -1, -1, 70.4137931034
+5288290650.jpg, -1, -1, 74.0818584071
+5292219744.jpg, -1, -1, 29.9115044248
+5298734322.jpg, -1, -1, 64.9779411765
+5300293470.jpg, -1, -1, 57.0848623853
+5306806108.jpg, -1, -1, 69.7731092437
+5309326426.jpg, -1, -1, 71.0884955752
+5314378959.jpg, -1, -1, 23.8684210526
+5314923965.jpg, -1, -1, 49.7289719626
+5320165275.jpg, -1, -1, 70.796460177
+5323106663.jpg, -1, -1, 54.4456521739
+5337507620.jpg, -1, -1, 63.2433035714
+5338193826.jpg, -1, -1, 54.1418918919
+5344616817.jpg, -1, -1, 58.5384615385
+5346690387.jpg, -1, -1, 51.9864864865
+5354068609.jpg, -1, -1, 62.9142857143
+5367389511.jpg, -1, -1, 58.8973214286
+5378235552.jpg, -1, -1, 67.8461538462
+5378725814.jpg, -1, -1, 57.3628318584
+5394502414.jpg, -1, -1, 46.3349514563
+5399180378.jpg, -1, -1, 69.8589108911
+5411857475.jpg, -1, -1, 58.0428571429
+5417875149.jpg, -1, -1, 39.0540540541
+5423246371.jpg, -1, -1, 68.3107476636
+5433512452.jpg, -1, -1, 61.5231092437
+5438123525.jpg, -1, -1, 41.2543103448
+5443343893.jpg, -1, -1, 59.265625
+5444850191.jpg, -1, -1, 67.233490566
+5448376373.jpg, -1, -1, 82.9551886792
+5449529114.jpg, -1, -1, 72.3804347826
+5455599246.jpg, -1, -1, 58.9813084112
+5462654627.jpg, -1, -1, 77.0434782609
+5467862729.jpg, -1, -1, 74.9579646018
+5474201544.jpg, -1, -1, 65.0630434783
+5476657075.jpg, -1, -1, 66.1674311927
+5480545191.jpg, -1, -1, 31.3205128205
+5480778835.jpg, -1, -1, 56.4621848739
+5481249377.jpg, -1, -1, 36.2878151261
+5483224233.jpg, -1, -1, 62.354368932
+5483514140.jpg, -1, -1, 74.7605932203
+5500570132.jpg, -1, -1, 63.9855769231
+5505340753.jpg, -1, -1, 46.4789915966
+5506716444.jpg, -1, -1, 61.775
+5507113852.jpg, -1, -1, 66.5147058824
+5525979481.jpg, -1, -1, 46.375
+5527745877.jpg, -1, -1, 84.4166666667
+5540702638.jpg, -1, -1, 72.7056074766
+5541438392.jpg, -1, -1, 45.7457627119
+5542509126.jpg, -1, -1, 73.9977678571
+5550230387.jpg, -1, -1, 62.0051020408
+5554551412.jpg, -1, -1, 32.1747787611
+5561903765.jpg, -1, -1, 56.141025641
+5562990409.jpg, -1, -1, 72.2369565217
+5565514955.jpg, -1, -1, 45.9587155963
+5570837854.jpg, -1, -1, 55.8296460177
+5571768533.jpg, -1, -1, 54.6961206897
+5583866722.jpg, -1, -1, 40.875
+5592031827.jpg, -1, -1, 55.3058252427
+5592471041.jpg, -1, -1, 66.2567567568
+5594332636.jpg, -1, -1, 69.6490825688
+5596136215.jpg, -1, -1, 57.3869565217
+5596270310.jpg, -1, -1, 60.8714285714
+5597328203.jpg, -1, -1, 48.5287610619
+5606263347.jpg, -1, -1, 21.1586956522
+5610005411.jpg, -1, -1, 69.0798319328
+5617865491.jpg, -1, -1, 53.9110576923
+5631740117.jpg, -1, -1, 77.2212389381
+563494328.jpg, -1, -1, 25.68125
+5643887824.jpg, -1, -1, 69.3040540541
+5644717504.jpg, -1, -1, 46.5575221239
+5651460875.jpg, -1, -1, 69.6517857143
+5656752407.jpg, -1, -1, 77.1869565217
+5657622409.jpg, -1, -1, 68.2452830189
+5660186741.jpg, -1, -1, 37.3970588235
+5669528587.jpg, -1, -1, 63.5428571429
+5679431462.jpg, -1, -1, 72.593220339
+5680314396.jpg, -1, -1, 48.4025423729
+5681801478.jpg, -1, -1, 46.6351351351
+5682574149.jpg, -1, -1, 51.3402777778
+5705716016.jpg, -1, -1, 38.8241525424
+5712901839.jpg, -1, -1, 50.1972477064
+5724255636.jpg, -1, -1, 44.6071428571
+5730689688.jpg, -1, -1, 72.268907563
+5739537678.jpg, -1, -1, 46.1282051282
+5740000928.jpg, -1, -1, 69.0105042017
+5765859594.jpg, -1, -1, 74.102173913
+5766806996.jpg, -1, -1, 23.61875
+5770080490.jpg, -1, -1, 65.6134453782
+5807318968.jpg, -1, -1, 51.5487288136
+5812840397.jpg, -1, -1, 70.9487179487
+5825508500.jpg, -1, -1, 21.8541666667
+5825939601.jpg, -1, -1, 78.7120535714
+5832751066.jpg, -1, -1, 73.7941176471
+5846011911.jpg, -1, -1, 36.8986486486
+5848411050.jpg, -1, -1, 79.4121621622
+5849249346.jpg, -1, -1, 77.4210526316
+5850096725.jpg, -1, -1, 36.7
+5866448510.jpg, -1, -1, 69.0810810811
+5869591890.jpg, -1, -1, 70.2860169492
+5870845920.jpg, -1, -1, 72.4742990654
+5876597412.jpg, -1, -1, 57.9676724138
+5880149356.jpg, -1, -1, 53.0972222222
+5889219072.jpg, -1, -1, 68.3871681416
+5890838382.jpg, -1, -1, 61.375
+5891304676.jpg, -1, -1, 53.2266949153
+5901783769.jpg, -1, -1, 63.7717391304
+5906582267.jpg, -1, -1, 70.975
+5915008700.jpg, -1, -1, 51.7635135135
+5919706356.jpg, -1, -1, 47.2876106195
+5920883206.jpg, -1, -1, 61.7433628319
+5922588235.jpg, -1, -1, 72.0657894737
+5923067948.jpg, -1, -1, 72.2629310345
+5932454750.jpg, -1, -1, 64.675
+5936466945.jpg, -1, -1, 52.1351351351
+5946379420.jpg, -1, -1, 54.4247572816
+5950051723.jpg, -1, -1, 75.4117647059
+5958950790.jpg, -1, -1, 33.3067226891
+5959668491.jpg, -1, -1, 72.4051724138
+5970029392.jpg, -1, -1, 64.1575630252
+5973304617.jpg, -1, -1, 65.9956521739
+5990072932.jpg, -1, -1, 61.7820512821
+5998921543.jpg, -1, -1, 31.5107758621
+6008267730.jpg, -1, -1, 31.25
+6026326925.jpg, -1, -1, 62.6323529412
+6031843630.jpg, -1, -1, 73.3781512605
+6035849303.jpg, -1, -1, 64.391025641
+6046652289.jpg, -1, -1, 73.0911214953
+6059676315.jpg, -1, -1, 69.972972973
+6066179576.jpg, -1, -1, 54.1907894737
+6067656779.jpg, -1, -1, 73.781779661
+6070993297.jpg, -1, -1, 57.3095238095
+6071930227.jpg, -1, -1, 21.14375
+6080237953.jpg, -1, -1, 59.5068807339
+6096713881.jpg, -1, -1, 40.4077669903
+6098070477.jpg, -1, -1, 64.9932432432
+6100280225.jpg, -1, -1, 55.2683486239
+6106078585.jpg, -1, -1, 65.0168269231
+6107409429.jpg, -1, -1, 53.1794871795
+6109808754.jpg, -1, -1, 75.8093220339
+6112607748.jpg, -1, -1, 72.1467889908
+6113677746.jpg, -1, -1, 75.5767326733
+6127342377.jpg, -1, -1, 35.147826087
+6143020134.jpg, -1, -1, 68.5277777778
+6162845121.jpg, -1, -1, 71.0871559633
+6163563949.jpg, -1, -1, 72.025
+6163576407.jpg, -1, -1, 63.5067567568
+6164643779.jpg, -1, -1, 68.5782608696
+6167276187.jpg, -1, -1, 16.2307692308
+6169751513.jpg, -1, -1, 68.125
+6172349199.jpg, -1, -1, 71.9915966387
+6175048748.jpg, -1, -1, 70.9040178571
+6175479501.jpg, -1, -1, 59.6752336449
+6184879643.jpg, -1, -1, 62.3175675676
+6190472459.jpg, -1, -1, 64.5608695652
+6201268474.jpg, -1, -1, 65.7889908257
+6201290598.jpg, -1, -1, 65.6621621622
+6205974624.jpg, -1, -1, 51.9558823529
+6222011944.jpg, -1, -1, 63.1305309735
+6222668667.jpg, -1, -1, 70.0
+6227646726.jpg, -1, -1, 64.027027027
+6228278450.jpg, -1, -1, 62.6239130435
+6230011878.jpg, -1, -1, 77.4797297297
+6238988012.jpg, -1, -1, 52.4913793103
+6244943274.jpg, -1, -1, 39.2629310345
+6252195920.jpg, -1, -1, 71.7542372881
+6252913864.jpg, -1, -1, 48.2760869565
+6255855320.jpg, -1, -1, 54.1330275229
+6256521905.jpg, -1, -1, 72.6816037736
+6256829619.jpg, -1, -1, 78.6507633588
+6258120264.jpg, -1, -1, 71.5995575221
+6272061393.jpg, -1, -1, 73.6891891892
+6274128274.jpg, -1, -1, 61.1173913043
+6286044979.jpg, -1, -1, 66.3060747664
+6309109435.jpg, -1, -1, 33.3944954128
+6321717266.jpg, -1, -1, 54.9806034483
+6337822588.jpg, -1, -1, 47.9105839416
+6351773844.jpg, -1, -1, 44.8151260504
+6354614287.jpg, -1, -1, 49.3318584071
+6366256547.jpg, -1, -1, 71.3075221239
+6379014005.jpg, -1, -1, 44.7544642857
+6379443681.jpg, -1, -1, 47.921875
+6389008737.jpg, -1, -1, 69.6462264151
+6439499095.jpg, -1, -1, 67.525
+6459542105.jpg, -1, -1, 55.75
+6468297341.jpg, -1, -1, 56.7923728814
+6469192803.jpg, -1, -1, 32.0848214286
+6486375983.jpg, -1, -1, 20.6612149533
+6504283859.jpg, -1, -1, 49.8944954128
+6534779805.jpg, -1, -1, 34.3101503759
+6534801331.jpg, -1, -1, 49.8653846154
+6537543865.jpg, -1, -1, 32.5398230088
+6551619395.jpg, -1, -1, 69.0442477876
+6558436183.jpg, -1, -1, 73.8119266055
+656718644.jpg, -1, -1, 27.9711538462
+6576363795.jpg, -1, -1, 31.7640186916
+6598327095.jpg, -1, -1, 62.0215517241
+6605703733.jpg, -1, -1, 71.3676470588
+6621080809.jpg, -1, -1, 47.9615384615
+665094052.jpg, -1, -1, 29.8026315789
+6656519345.jpg, -1, -1, 26.3392857143
+6686438027.jpg, -1, -1, 69.175
+6699302891.jpg, -1, -1, 81.7162162162
+6703759327.jpg, -1, -1, 66.7642857143
+6711319143.jpg, -1, -1, 74.5625
+6749061363.jpg, -1, -1, 64.6201923077
+6790082633.jpg, -1, -1, 56.222972973
+6795087227.jpg, -1, -1, 29.39375
+6803864259.jpg, -1, -1, 62.6
+6810784384.jpg, -1, -1, 47.1554054054
+6812098169.jpg, -1, -1, 75.628440367
+6831984052.jpg, -1, -1, 76.3014705882
+6835820588.jpg, -1, -1, 59.4673913043
+6838263860.jpg, -1, -1, 39.6896551724
+6840130023.jpg, -1, -1, 71.3043478261
+6849285524.jpg, -1, -1, 27.275210084
+6849557935.jpg, -1, -1, 23.0723684211
+6850513360.jpg, -1, -1, 78.8913793103
+6854146276.jpg, -1, -1, 73.1081730769
+6865288108.jpg, -1, -1, 69.6776315789
+6873503299.jpg, -1, -1, 44.7179487179
+6875112664.jpg, -1, -1, 62.7073170732
+6896225625.jpg, -1, -1, 62.6323529412
+6897675527.jpg, -1, -1, 56.9935483871
+6898490953.jpg, -1, -1, 80.1554054054
+6904500540.jpg, -1, -1, 74.8057692308
+6905574433.jpg, -1, -1, 44.2775423729
+6907499876.jpg, -1, -1, 70.3883928571
+6908276949.jpg, -1, -1, 66.7226890756
+6922435153.jpg, -1, -1, 66.2142857143
+6935219850.jpg, -1, -1, 45.186440678
+6935652834.jpg, -1, -1, 35.2375
+6943168420.jpg, -1, -1, 61.6964285714
+6953798231.jpg, -1, -1, 48.3411214953
+6959482720.jpg, -1, -1, 75.8739495798
+6961812492.jpg, -1, -1, 63.5263157895
+6979435146.jpg, -1, -1, 81.1637168142
+6980277214.jpg, -1, -1, 61.3990825688
+6989772902.jpg, -1, -1, 12.7457627119
+6993179262.jpg, -1, -1, 35.2635135135
+7002011797.jpg, -1, -1, 70.2839805825
+7004025487.jpg, -1, -1, 24.375
+7004286982.jpg, -1, -1, 77.7116935484
+7006133071.jpg, -1, -1, 70.5940594059
+7017996705.jpg, -1, -1, 59.0913793103
+7019640975.jpg, -1, -1, 76.2205882353
+7035194341.jpg, -1, -1, 76.8851351351
+7076980511.jpg, -1, -1, 62.3550420168
+7077356639.jpg, -1, -1, 71.0884955752
+7084961757.jpg, -1, -1, 64.202173913
+7094182327.jpg, -1, -1, 51.6092436975
+7100857351.jpg, -1, -1, 77.195754717
+7102405013.jpg, -1, -1, 64.1776315789
+7106381255.jpg, -1, -1, 6.61853448276
+7108939139.jpg, -1, -1, 81.1228813559
+7143312581.jpg, -1, -1, 56.8577981651
+7155826261.jpg, -1, -1, 69.2834821429
+7161322897.jpg, -1, -1, 66.6144859813
+7165207702.jpg, -1, -1, 67.5945945946
+7167608916.jpg, -1, -1, 66.4054054054
+7169475537.jpg, -1, -1, 82.6189320388
+7173633933.jpg, -1, -1, 34.5689655172
+7194820358.jpg, -1, -1, 80.8716814159
+7200434768.jpg, -1, -1, 64.9530075188
+7203007382.jpg, -1, -1, 38.0526315789
+7214629206.jpg, -1, -1, 68.6176470588
+7219282312.jpg, -1, -1, 55.4639830508
+7238585598.jpg, -1, -1, 66.4889380531
+7247521810.jpg, -1, -1, 34.34375
+7250719968.jpg, -1, -1, 76.4514563107
+7255289130.jpg, -1, -1, 74.5986842105
+7260432838.jpg, -1, -1, 69.963592233
+7263783306.jpg, -1, -1, 40.0401785714
+7264968696.jpg, -1, -1, 70.8136792453
+7289957138.jpg, -1, -1, 56.6875
+7304709380.jpg, -1, -1, 77.3514150943
+7321293264.jpg, -1, -1, 49.0775862069
+7339222496.jpg, -1, -1, 65.125
+7348931184.jpg, -1, -1, 58.9333333333
+7358753494.jpg, -1, -1, 54.1504424779
+7361922724.jpg, -1, -1, 78.4390756303
+7406520610.jpg, -1, -1, 81.2847222222
+7411918198.jpg, -1, -1, 49.311440678
+7418763178.jpg, -1, -1, 55.0571428571
+7433488844.jpg, -1, -1, 48.6666666667
+7442324444.jpg, -1, -1, 59.7162162162
+7459539092.jpg, -1, -1, 64.9736842105
+7459547170.jpg, -1, -1, 65.1584821429
+7470463166.jpg, -1, -1, 73.1485849057
+7478826230.jpg, -1, -1, 67.66
+7502944264.jpg, -1, -1, 72.01875
+7512526314.jpg, -1, -1, 62.0084033613
+7526349354.jpg, -1, -1, 36.875
+7555167080.jpg, -1, -1, 64.2268907563
+7570426860.jpg, -1, -1, 73.0127118644
+7597470314.jpg, -1, -1, 52.9308035714
+7627158600.jpg, -1, -1, 61.4485981308
+7628564828.jpg, -1, -1, 67.2894736842
+7628769918.jpg, -1, -1, 76.2430555556
+7629156184.jpg, -1, -1, 70.375
+7640299878.jpg, -1, -1, 68.5608108108
+7641363622.jpg, -1, -1, 66.0833333333
+764877350.jpg, -1, -1, 52.975
+7651161432.jpg, -1, -1, 80.2146017699
+7664231886.jpg, -1, -1, 67.5202702703
+7668828488.jpg, -1, -1, 70.675
+7678821440.jpg, -1, -1, 73.3088235294
+767916820.jpg, -1, -1, 71.2867647059
+7685277932.jpg, -1, -1, 62.5
+7714653298.jpg, -1, -1, 46.0066964286
+772323238.jpg, -1, -1, 47.0183486239
+7739550292.jpg, -1, -1, 56.275
+7755789428.jpg, -1, -1, 63.5754716981
+7786382886.jpg, -1, -1, 59.6752336449
+7795515978.jpg, -1, -1, 55.0833333333
+7811282766.jpg, -1, -1, 71.2364864865
+7820783548.jpg, -1, -1, 42.3256880734
+7822391242.jpg, -1, -1, 60.0441176471
+7823939538.jpg, -1, -1, 74.8913043478
+7850123296.jpg, -1, -1, 54.4322429907
+7856451374.jpg, -1, -1, 68.875
+7859713146.jpg, -1, -1, 77.5572033898
+7866480612.jpg, -1, -1, 72.4481132075
+7871237210.jpg, -1, -1, 73.8152173913
+7883125264.jpg, -1, -1, 45.4410377358
+7904631250.jpg, -1, -1, 76.8012820513
+7914158128.jpg, -1, -1, 65.0460526316
+7920824878.jpg, -1, -1, 62.5
+7926649668.jpg, -1, -1, 47.5649038462
+7935628474.jpg, -1, -1, 77.1377118644
+7940282852.jpg, -1, -1, 43.4285714286
+7946535104.jpg, -1, -1, 63.4642857143
+7954181598.jpg, -1, -1, 64.8638392857
+7957627098.jpg, -1, -1, 38.4745762712
+7975750167.jpg, -1, -1, 71.640625
+7983388187.jpg, -1, -1, 74.8340336134
+7989520325.jpg, -1, -1, 50.0105932203
+7994018610.jpg, -1, -1, 72.2436440678
+8002007313.jpg, -1, -1, 75.1714285714
+8006234274.jpg, -1, -1, 76.3451327434
+8010069396.jpg, -1, -1, 73.8275862069
+8012731043.jpg, -1, -1, 60.0783898305
+8034708472.jpg, -1, -1, 62.8177083333
+8044706176.jpg, -1, -1, 39.0769230769
+8060595540.jpg, -1, -1, 65.1584821429
+8077634378.jpg, -1, -1, 68.44375
+8081563599.jpg, -1, -1, 35.8857142857
+8103481984.jpg, -1, -1, 47.8
+8103867208.jpg, -1, -1, 70.0200892857
+8104089685.jpg, -1, -1, 60.2364864865
+8109380358.jpg, -1, -1, 47.5487804878
+8113245382.jpg, -1, -1, 79.03125
+8127846229.jpg, -1, -1, 46.1540178571
+8130745750.jpg, -1, -1, 57.3221153846
+8132171428.jpg, -1, -1, 67.4459459459
+8135725807.jpg, -1, -1, 71.125
+8137767984.jpg, -1, -1, 14.261682243
+8142555462.jpg, -1, -1, 47.6736111111
+8152535252.jpg, -1, -1, 17.8027522936
+8157884969.jpg, -1, -1, 45.328358209
+8167934912.jpg, -1, -1, 74.6153846154
+8175674967.jpg, -1, -1, 69.9413043478
+8182062575.jpg, -1, -1, 43.1108695652
+8199832249.jpg, -1, -1, 61.988317757
+8228733135.jpg, -1, -1, 59.9181415929
+8229993789.jpg, -1, -1, 63.8040540541
+8236704145.jpg, -1, -1, 72.6793478261
+8240021404.jpg, -1, -1, 58.9690265487
+8240220210.jpg, -1, -1, 45.231092437
+8242599767.jpg, -1, -1, 72.0089285714
+8246007228.jpg, -1, -1, 51.4252336449
+8247844062.jpg, -1, -1, 66.6886792453
+8257636356.jpg, -1, -1, 71.9345794393
+8259557757.jpg, -1, -1, 63.3509615385
+8262794558.jpg, -1, -1, 44.5334821429
+8278414113.jpg, -1, -1, 58.31
+8279922752.jpg, -1, -1, 73.975
+8293206684.jpg, -1, -1, 66.1674311927
+8293797976.jpg, -1, -1, 58.2740384615
+8301114122.jpg, -1, -1, 53.8673469388
+8305612260.jpg, -1, -1, 60.3066037736
+8331607175.jpg, -1, -1, 70.7754237288
+8333619480.jpg, -1, -1, 52.4369565217
+8339157161.jpg, -1, -1, 65.9778761062
+8345197886.jpg, -1, -1, 64.1776315789
+8345496530.jpg, -1, -1, 72.8716216216
+8350531104.jpg, -1, -1, 70.6
+8353492181.jpg, -1, -1, 29.9449152542
+8364603294.jpg, -1, -1, 80.0565217391
+8381849363.jpg, -1, -1, 72.8525641026
+8385499383.jpg, -1, -1, 69.5631067961
+8385552455.jpg, -1, -1, 77.0915178571
+8391472456.jpg, -1, -1, 66.5619469027
+8406395791.jpg, -1, -1, 69.0442477876
+8410894233.jpg, -1, -1, 47.9926470588
+8412239545.jpg, -1, -1, 85.5275423729
+8412460969.jpg, -1, -1, 38.3421052632
+8417110649.jpg, -1, -1, 48.017699115
+8421578109.jpg, -1, -1, 61.4766949153
+8428610140.jpg, -1, -1, 68.1
+8430005179.jpg, -1, -1, 71.3948598131
+8430084372.jpg, -1, -1, 50.0995145631
+8431773301.jpg, -1, -1, 66.9263392857
+8439178228.jpg, -1, -1, 46.7569444444
+8441623412.jpg, -1, -1, 75.8289473684
+8443583255.jpg, -1, -1, 72.8108695652
+8451346661.jpg, -1, -1, 76.7881355932
+8458755761.jpg, -1, -1, 70.349009901
+8460289501.jpg, -1, -1, 65.0287610619
+8461148661.jpg, -1, -1, 47.0183486239
+8466537488.jpg, -1, -1, 74.5199115044
+8468743988.jpg, -1, -1, 77.9256756757
+8469388516.jpg, -1, -1, 58.6039823009
+8481265200.jpg, -1, -1, 70.3970588235
+8481724236.jpg, -1, -1, 42.0384615385
+8484786490.jpg, -1, -1, 67.4729299363
+8488084256.jpg, -1, -1, 68.44375
+8502941912.jpg, -1, -1, 53.8154205607
+8506941154.jpg, -1, -1, 58.3881578947
+8522094125.jpg, -1, -1, 65.8676470588
+8532549490.jpg, -1, -1, 66.9306722689
+8539118741.jpg, -1, -1, 52.9264705882
+8544556866.jpg, -1, -1, 65.453125
+8546536299.jpg, -1, -1, 70.575
+8551620615.jpg, -1, -1, 70.9040178571
+8552101917.jpg, -1, -1, 32.5486111111
+8553130419.jpg, -1, -1, 62.8378378378
+8553449461.jpg, -1, -1, 71.2307692308
+8554330315.jpg, -1, -1, 74.3738938053
+8556290824.jpg, -1, -1, 67.5606796117
+8560249157.jpg, -1, -1, 33.93125
+8572309728.jpg, -1, -1, 74.34765625
+8573294301.jpg, -1, -1, 33.6764705882
+8585320404.jpg, -1, -1, 66.8539823009
+8595112756.jpg, -1, -1, 77.5862831858
+8595845896.jpg, -1, -1, 57.4864864865
+8596148572.jpg, -1, -1, 61.75
+8596349698.jpg, -1, -1, 78.25
+8599488070.jpg, -1, -1, 55.4797297297
+8599891034.jpg, -1, -1, 62.7302631579
+8608196738.jpg, -1, -1, 65.6514423077
+8608242752.jpg, -1, -1, 75.0957943925
+8613815314.jpg, -1, -1, 63.0847457627
+8616708613.jpg, -1, -1, 32.4618644068
+8617343673.jpg, -1, -1, 76.4285714286
+8617452211.jpg, -1, -1, 62.9975247525
+86243803.jpg, -1, -1, 39.01875
+8626945283.jpg, -1, -1, 56.9407894737
+8637618858.jpg, -1, -1, 73.2071428571
+8639628841.jpg, -1, -1, 71.1949152542
+8642423630.jpg, -1, -1, 66.4210526316
+8647880189.jpg, -1, -1, 69.0986842105
+8652814108.jpg, -1, -1, 40.9
+8653149698.jpg, -1, -1, 62.53125
+8654833683.jpg, -1, -1, 62.8371559633
+8665123695.jpg, -1, -1, 64.9724576271
+8671346109.jpg, -1, -1, 59.4866071429
+8688672632.jpg, -1, -1, 77.1377118644
+8703113665.jpg, -1, -1, 46.3026315789
+8718552383.jpg, -1, -1, 73.8504464286
+8728473663.jpg, -1, -1, 66.8526785714
+8731143356.jpg, -1, -1, 65.4285714286
+8768474112.jpg, -1, -1, 27.2435344828
+8771756230.jpg, -1, -1, 22.1725663717
+8774523686.jpg, -1, -1, 72.5233050847
+8779946210.jpg, -1, -1, 76.6880733945
+8816701721.jpg, -1, -1, 49.7176724138
+8836006786.jpg, -1, -1, 73.785046729
+8841877925.jpg, -1, -1, 74.7521551724
+8853329035.jpg, -1, -1, 66.2826086957
+8871477031.jpg, -1, -1, 15.8220338983
+8877350801.jpg, -1, -1, 44.369266055
+8892359151.jpg, -1, -1, 55.1824324324
+8897373737.jpg, -1, -1, 72.0338983051
+8905578199.jpg, -1, -1, 77.1482300885
+8911411937.jpg, -1, -1, 75.4879807692
+8916833328.jpg, -1, -1, 56.5452586207
+8924220861.jpg, -1, -1, 74.6810344828
+8925875609.jpg, -1, -1, 76.8066037736
+8929337311.jpg, -1, -1, 59.6582568807
+8939732555.jpg, -1, -1, 17.7152173913
+8942422332.jpg, -1, -1, 72.6009174312
+8944011983.jpg, -1, -1, 29.4344660194
+8971376273.jpg, -1, -1, 79.0405405405
+8972898508.jpg, -1, -1, 28.01875
+8973350530.jpg, -1, -1, 72.4
+898061275.jpg, -1, -1, 75.4347014925
+9016868192.jpg, -1, -1, 68.1785714286
+9018209858.jpg, -1, -1, 43.4487179487
+9037452749.jpg, -1, -1, 34.0723684211
+9053743376.jpg, -1, -1, 68.1785714286
+9054377711.jpg, -1, -1, 72.2146226415
+9061831747.jpg, -1, -1, 38.5490654206
+9063333592.jpg, -1, -1, 44.3125
+9111487897.jpg, -1, -1, 75.1782608696
+9125517985.jpg, -1, -1, 47.0558252427
+9130908149.jpg, -1, -1, 63.5336134454
+9132059359.jpg, -1, -1, 55.2142857143
+9136502604.jpg, -1, -1, 73.031512605
+9142684025.jpg, -1, -1, 73.5861344538
+9146239672.jpg, -1, -1, 65.00625
+9166911645.jpg, -1, -1, 61.4294871795
+9201324227.jpg, -1, -1, 67.9575892857
+9211639075.jpg, -1, -1, 79.8581081081
+9215122110.jpg, -1, -1, 44.35
+9258012910.jpg, -1, -1, 76.2987288136
+9290273110.jpg, -1, -1, 47.921875
+9306814074.jpg, -1, -1, 74.9004237288
+9336450850.jpg, -1, -1, 79.5172413793
+9339437714.jpg, -1, -1, 64.9375
+9346802331.jpg, -1, -1, 43.41875
+935105924.jpg, -1, -1, 50.2441860465
+9369983914.jpg, -1, -1, 70.4551282051
+9380189805.jpg, -1, -1, 70.3785714286
+9407500995.jpg, -1, -1, 75.8792372881
+9418862819.jpg, -1, -1, 32.8813559322
+9419143176.jpg, -1, -1, 48.25
+9428417271.jpg, -1, -1, 69.325
+9445529272.jpg, -1, -1, 44.2388392857
+9461924966.jpg, -1, -1, 59.265625
+9463299362.jpg, -1, -1, 61.6486486486
+9480350471.jpg, -1, -1, 59.9399038462
+9482188665.jpg, -1, -1, 69.0263157895
+9482512663.jpg, -1, -1, 80.4146341463
+9484990711.jpg, -1, -1, 30.7
+9485553260.jpg, -1, -1, 77.8990825688
+95019293.jpg, -1, -1, 37.3970588235
+9513920516.jpg, -1, -1, 55.3241525424
+9526812010.jpg, -1, -1, 71.8963414634
+9527733918.jpg, -1, -1, 67.8175675676
+9557331185.jpg, -1, -1, 53.0553097345
+9560250736.jpg, -1, -1, 43.8717948718
+9569803128.jpg, -1, -1, 69.8473451327
+9583278733.jpg, -1, -1, 74.60546875
+9583424619.jpg, -1, -1, 62.8378378378
+9605307671.jpg, -1, -1, 68.6205357143
+9619949433.jpg, -1, -1, 47.6736111111
+9623701556.jpg, -1, -1, 72.3296460177
+9629668489.jpg, -1, -1, 74.4097222222
+9646156379.jpg, -1, -1, 71.8440366972
+9654367147.jpg, -1, -1, 70.8823529412
+9677256773.jpg, -1, -1, 30.6415929204
+9678381010.jpg, -1, -1, 51.237804878
+9686337427.jpg, -1, -1, 29.9142857143
+9699196303.jpg, -1, -1, 72.1043165468
+9723836931.jpg, -1, -1, 79.1747572816
+9768810935.jpg, -1, -1, 55.2567567568
+9784161203.jpg, -1, -1, 20.25
+9795395745.jpg, -1, -1, 60.1378504673
+9823108104.jpg, -1, -1, 69.9027777778
+9826425224.jpg, -1, -1, 64.1526548673
+9831633846.jpg, -1, -1, 65.1584821429
+9849484955.jpg, -1, -1, 45.4910714286
+9895678845.jpg, -1, -1, 57.7476635514
+9941134286.jpg, -1, -1, 65.9048672566
+9945099433.jpg, -1, -1, 50.4329268293
+9954707815.jpg, -1, -1, 56.1631355932
+9960023395.jpg, -1, -1, 65.474789916

examplar_data_labels/LIVE_Qualcomm/labels.txt

@@ -0,0 +1,208 @@
+Focus/1024_TableTennisAtBalboa_HTCOneVx_VIDEO0138.yuv, -1, -1, 41.64888182462034
+Focus/1019_SkateBoardingSkills_LGG2_CAM00855.yuv, -1, -1, 54.094978137969605
+Focus/1019_SkateBoardingSkills_OppoFind7_VID20150507134738.yuv, -1, -1, 51.02291495174518
+Exposure/0911panGarden_Note4_20150911_152420.yuv, -1, -1, 44.69628201977274
+Artifacts/0923_ChattingOverLunch_G2_CAM00813.yuv, -1, -1, 45.67855937078179
+Artifacts/0911nightScene_note4_20150911_193809.yuv, -1, -1, 36.546723521311996
+Exposure/0912_Burgers_Nokia1020_20130306_17_07_38_Pro.yuv, -1, -1, 64.9396364629356
+Color/1006_EmergencyScene_GS6_20151006_160038.yuv, -1, -1, 58.33545178737855
+Stabilization/0923_WalkingToLunch2_G2_CAM00812.yuv, -1, -1, 48.7769838458931
+Focus/1006_ComplexTrain_HTCOneVX_VIDEO0087.yuv, -1, -1, 22.46553753382192
+Color/0913playSand_GS5_20150913_183702.yuv, -1, -1, 58.51617627453426
+Stabilization/1019_ConstructionBuggy2_GS6_20151020_133229.yuv, -1, -1, 61.262155162968625
+Sharpness/0913taiZi_iPhone5s_IMG_0250.yuv, -1, -1, 29.54546786777324
+Artifacts/1006_FollowTheTrain_OppoFind7_VID20150424111801.yuv, -1, -1, 48.7312983952301
+Stabilization/1019_ConstructionBuggy2_LGG2_CAM00867.yuv, -1, -1, 42.73349382920229
+Artifacts/0918_AxelFetchesBall_GS6_20150919_090332.yuv, -1, -1, 58.23142265391711
+Exposure/1019_ViewFromMetro2_OppoFind7_VID20150508085916.yuv, -1, -1, 36.18976114116138
+Exposure/0923_FollowTheBee_HTCOneVx_VIDEO0072.yuv, -1, -1, 33.04439956164501
+Artifacts/1019_ColorfulKites2_LGG2_CAM00872.yuv, -1, -1, 63.17108813219617
+Exposure/0923_FollowTheBee_GS6_20150923_131355.yuv, -1, -1, 51.45085729576881
+Color/1024_BadmintonAtBalboa_GS6_20151024_104414.yuv, -1, -1, 56.589364068655456
+Exposure/1022_QCOMSoccer2_HTCOneVx_VIDEO0137.yuv, -1, -1, 29.11684686808589
+Sharpness/0912_football_Nokia1020_20130306_17_51_35_Pro.yuv, -1, -1, 57.57359450935232
+Artifacts/0723_ManUnderTree_GS6_03_20150723_130100.yuv, -1, -1, 57.28880925319935
+Exposure/1019_ViewFromMetro2_GS6_20151020_133722.yuv, -1, -1, 39.88244836115969
+Exposure/1019_TrafficFromAfar_OppoFind7_VID20150505125608.yuv, -1, -1, 43.74671562206279
+Artifacts/0918_AxelFetchesBall_HTCOneVX_VIDEO0056.yuv, -1, -1, 42.628796648669706
+Stabilization/0923_ChasingDucksInWater_iphone5_IMG_0274.yuv, -1, -1, 51.11518283501527
+Focus/1006_ComplexTrain_GS6_20151006_155711.yuv, -1, -1, 44.35577617160543
+Artifacts/1006_DucksAndFish2_HTCOneVX_VIDEO0093.yuv, -1, -1, 31.435259054533738
+Color/1024_BadmintonAtBalboa_HTCOneVx_VIDEO0139.yuv, -1, -1, 38.4966864084717
+Artifacts/1006_FollowTheTrain_GS6_20151006_155539.yuv, -1, -1, 47.30872740564676
+Sharpness/0912_football_GS5_20150912_153941.yuv, -1, -1, 53.23486148188893
+Exposure/0911_DogBall_OppoFind7_VID20150330141759.yuv, -1, -1, 66.94118272054322
+Focus/0923_BirdsFlyOnPier_iphone5_IMG_0277.yuv, -1, -1, 24.644612569329546
+Stabilization/1019_MetroArrives_OppoFind7_VID20150508085607.yuv, -1, -1, 55.86662414448337
+Exposure/0911panGarden_Lumia1020_WindowsPhone_20130330_23_52_26_Pro.yuv, -1, -1, 61.25422142212514
+Sharpness/1006_ParkBus_OppoFind7_VID20150424122747.yuv, -1, -1, 55.317108648036935
+Stabilization/1019_ConstructionBuggy2_HTCOneVX_VIDEO0119.yuv, -1, -1, 37.69768941262607
+Sharpness/0918_GuysPlayVolleyball_GS6_20150919_093108.yuv, -1, -1, 65.25654758810917
+Sharpness/1006_ParkBus_GS6_20151006_170532.yuv, -1, -1, 61.532520666710234
+Artifacts/0918_AxelFetchesBall_LGG2_CAM00802.yuv, -1, -1, 52.891660679374006
+Focus/1019_FlyingKite_LGG2_CAM00875.yuv, -1, -1, 42.076217605884914
+Color/1006_FamilyScooter_HTCOneVX_VIDEO0101.yuv, -1, -1, 38.8106247696991
+Color/1006_EmergencyScene_OppoFind7_VID20150424112300.yuv, -1, -1, 62.01071121403491
+Artifacts/0911nightScene_gs5_20150911_194737.yuv, -1, -1, 32.27546017222677
+Exposure/0911panGarden_iPhone5s_IMG_0132.yuv, -1, -1, 56.68920900925609
+Sharpness/0911duckSwim_iphone5s_IMG_0136.yuv, -1, -1, 69.15445557891265
+Stabilization/0923_SwingingArm_Note4_20150925_153938.yuv, -1, -1, 48.1232752921398
+Sharpness/0913taiZi_GS5_20150913_183616.yuv, -1, -1, 32.896692696712435
+Exposure/0911_DogBall_LGG2_CAM00791.yuv, -1, -1, 46.17375037863661
+Exposure/1022_QCOMSoccer2_LGG2_CAM00886.yuv, -1, -1, 26.395621305018185
+Focus/0913boyWalkBy_GS5_20150913_182938.yuv, -1, -1, 51.591805400221055
+Sharpness/0918_GuysPlayVolleyball_HTCOneVX_VIDEO0058.yuv, -1, -1, 53.13309561506308
+Sharpness/0923_VerticalAndRollerCoaster_iphone5_IMG_0258.yuv, -1, -1, 48.66419200608179
+Stabilization/1019_BollywoodDanceTraining_GS6_20151017_175412.yuv, -1, -1, 53.770511195731956
+Exposure/0817_WalkIntoBarn_iPhone5_IMG_0410.yuv, -1, -1, 39.516765825233485
+Color/1006_EmergencyScene_HTCOneVX_VIDEO0090.yuv, -1, -1, 42.55847413392969
+Artifacts/0923_ChattingOverLunch_GS6_20150923_122025.yuv, -1, -1, 60.43118323177181
+Color/0918_CyclistsNear1_OppoFind7_VID20150407033615.yuv, -1, -1, 50.22621242945224
+Stabilization/1019_SkateBoarderStepsOut_OppoFind7_VID20150507134909.yuv, -1, -1, 40.40425472582464
+Stabilization/0923_FollowPigeons_iphone5_IMG_0279.yuv, -1, -1, 47.7605951367109
+Sharpness/0826_TrainPullsAway_OppoFind7_VID20150727213123.yuv, -1, -1, 58.976466922446484
+Artifacts/0723_ManUnderTree_GS5_03_20150723_130016.yuv, -1, -1, 52.56657483287041
+Artifacts/0918_DogsOnBeach_HTCOneVX_VIDEO0055.yuv, -1, -1, 50.378497196338884
+Artifacts/0723_ManUnderTree_HTCOneVx_03_VIDEO0006.yuv, -1, -1, 36.709288593516966
+Color/0912_CheerLeader1_Nokia1080_20130306_17_46_21_Pro.yuv, -1, -1, 41.529339991183875
+Sharpness/0923_VerticalAndRollerCoaster_Note4_20150925_151648.yuv, -1, -1, 55.628924969845
+Color/0923_RiverBed_GS5_20150925_165808.yuv, -1, -1, 61.69082350125231
+Focus/1024_TableTennisAtBalboa_OppoFind7_VID20150512060301.yuv, -1, -1, 50.92808485111922
+Stabilization/0923_FollowPigeons_Nokia1020_WP_20130305_20_16_29_Pro.yuv, -1, -1, 52.181316514970064
+Color/1019_FlyingKite3_LGG2_CAM00877.yuv, -1, -1, 55.74782132350435
+Color/0918_CyclistsNear1_HTCOneVX_VIDEO0050.yuv, -1, -1, 43.90725048088645
+Color/1006_FamilyScooter_OppoFind7_VID20150424122855.yuv, -1, -1, 51.17307547834518
+Artifacts/0911nightScene_1020_WindowsPhone_20130305_21_59_44_Pro.yuv, -1, -1, 21.187371167047402
+Sharpness/0826_TrainPullsAway_iPhone5s_IMG_0423.yuv, -1, -1, 60.880133455554756
+Stabilization/0923_CrazySubmarine_Note4_20150925_160005.yuv, -1, -1, 56.08844663824663
+Focus/0923_BirdsFlyOnPier_Note4_20150925_164318.yuv, -1, -1, 34.77241986411562
+Artifacts/1006_DucksAndFish2_GS6_20151006_161642.yuv, -1, -1, 51.0393821878597
+Stabilization/1019_SkateBoarderStepsOut_LGG2_CAM00857.yuv, -1, -1, 48.672533412525446
+Exposure/0911_DogBall_HTCOneVx_VIDEO0045.yuv, -1, -1, 49.7123541769036
+Color/1019_FlyingKite3_HTCOneVX_VIDEO0129.yuv, -1, -1, 49.33731661081409
+Focus/1006_ComplexTrain_OppoFind7_VID20150424111934.yuv, -1, -1, 31.508494371587865
+Focus/0723_OldTownShop_GS6_12_20150723_133507.yuv, -1, -1, 64.9917581345629
+Stabilization/0923_ChasingDucksInWater_GS5_20150925_163521.yuv, -1, -1, 48.91083829259606
+Focus/0723_OldTownShop_Nokia1020_12_WP_20130426_22_46_05_Pro.yuv, -1, -1, 52.969252792370625
+Focus/1019_FlyingKite_HTCOneVX_VIDEO0127.yuv, -1, -1, 20.802340977962967
+Color/1006_FamilyScooter_LGG2_CAM00849.yuv, -1, -1, 49.68892718051139
+Stabilization/0923_ChasingDucksInWater_Nokia1020_WP_20130305_19_56_39_Pro.yuv, -1, -1, 40.700157633712266
+Sharpness/0827_OldTownFountain_GS5_20150827_180240.yuv, -1, -1, 66.22997014329692
+Exposure/0912_Burgers_GS5_20150912_145544.yuv, -1, -1, 63.475870057217286
+Color/1006_FamilyScooter_GS6_20151006_170632.yuv, -1, -1, 60.28464598447542
+Color/1024_BadmintonAtBalboa_LGG2_CAM00888.yuv, -1, -1, 52.3694844910836
+Stabilization/1019_BollywoodDanceTraining_OppoFind7_VID20150505131616.yuv, -1, -1, 49.18933619585497
+Exposure/0912_Burgers_Note4_20150912_144558.yuv, -1, -1, 61.98320494013754
+Stabilization/0923_WalkingToLunch2_GS6_20150923_120934.yuv, -1, -1, 46.50669762802673
+Stabilization/1019_MetroArrives_LGG2_CAM00868.yuv, -1, -1, 59.05090948407636
+Artifacts/1019_ColorfulKites2_GS6_20151020_140327.yuv, -1, -1, 69.34239333030179
+Focus/0923_BirdsFlyOnPier_GS5_20150925_165328.yuv, -1, -1, 46.858952126850156
+Focus/0923_BirdsFlyOnPier_Nokia1020_WP_20130305_20_14_47_Pro.yuv, -1, -1, 49.44336625534995
+Color/0923_RiverBed_iphone5_IMG_0280.yuv, -1, -1, 63.16961373461168
+Sharpness/0911duckSwim_gs5_20150911_154802.yuv, -1, -1, 59.46925058712509
+Color/0918_DelMarView_HTCOneVX_VIDEO0062.yuv, -1, -1, 45.79645003864048
+Color/0918_CyclistsNear1_LGG2_CAM00795.yuv, -1, -1, 46.753110388198024
+Exposure/0817_WalkIntoBarn_OppoFind7_VID20150718144333.yuv, -1, -1, 49.02405995062579
+Color/1019_FlyingKite3_GS6_20151020_141052.yuv, -1, -1, 73.64276415152705
+Artifacts/0918_DogsOnBeach_GS6_20150919_085853.yuv, -1, -1, 65.94233767581703
+Stabilization/0923_CrazySubmarine_Nokia1020_WP_20130305_19_31_30_Pro.yuv, -1, -1, 59.038670696087216
+Focus/0913boyWalkBy_iPhone5s_IMG_0245.yuv, -1, -1, 62.812549213807614
+Stabilization/1019_MetroArrives_GS6_20151020_133409.yuv, -1, -1, 65.82583000957678
+Sharpness/0911duckSwim_notes4_20150911_153835.yuv, -1, -1, 71.74875099099955
+Sharpness/1006_ParkBus_HTCOneVX_VIDEO0100.yuv, -1, -1, 31.50387615900128
+Focus/1019_SkateBoardingSkills_GS6_20151019_182541.yuv, -1, -1, 59.007584920405066
+Color/0918_DelMarView_LGG2_CAM00809.yuv, -1, -1, 48.7408776505235
+Color/0912_Cheerleader1_Note4_20150912_152444.yuv, -1, -1, 46.853275685229434
+Focus/1019_FlyingKite_OppoFind7_VID20150508093000.yuv, -1, -1, 54.075938062059876
+Exposure/1019_TrafficFromAfar_HTCOneVX_VIDEO0102.yuv, -1, -1, 38.676262337846296
+Focus/0911driving_5s_IMG_0145.yuv, -1, -1, 44.9586564442459
+Exposure/1019_ViewFromMetro2_LGG2_CAM00870.yuv, -1, -1, 37.58488437787315
+Focus/0723_OldTownShop_HTCOneVx_12_VIDEO0019.yuv, -1, -1, 30.53679255118817
+Stabilization/0923_ChasingDucksInWater_Note4_20150925_162515.yuv, -1, -1, 48.75250168989043
+Sharpness/0826_TrainPullsAway_Note4_20150826_175232.yuv, -1, -1, 57.182744558097575
+Focus/1019_FlyingKite_GS6_20151020_140804.yuv, -1, -1, 23.50289297893303
+Color/0918_CyclistsNear1_GS6_20150919_081325.yuv, -1, -1, 52.81500748332389
+Artifacts/1006_PlaneLandsAfar2_OppoFind7_VID20150424122651.yuv, -1, -1, 57.22288718627244
+Exposure/0912_Burgers_iphone5s_IMG_0167.yuv, -1, -1, 64.38345299460406
+Sharpness/0923_VerticalAndRollerCoaster_GS5_20150925_152655.yuv, -1, -1, 55.74790371519573
+Stabilization/0923_FollowPigeons_GS5_20150925_165510.yuv, -1, -1, 55.46436837100166
+Exposure/0817_WalkIntoBarn_LGG2_CAM00718.yuv, -1, -1, 51.146870112855
+Color/0913playSand_iPhone5s_IMG_0251.yuv, -1, -1, 72.96615929361013
+Stabilization/0923_CrazySubmarine_GS5_20150925_161013.yuv, -1, -1, 62.68393406357753
+Sharpness/0912_football_Note4_20150912_152954.yuv, -1, -1, 55.629277646191774
+Color/1006_EmergencyScene_LGG2_CAM00838.yuv, -1, -1, 66.8679745839632
+Stabilization/0923_CrazySubmarine_iphone5_IMG_0268.yuv, -1, -1, 62.539356287114956
+Color/0912_Cheerleader1_iphone_IMG_0184.yuv, -1, -1, 59.46973486351669
+Exposure/1019_TrafficFromAfar_GS6_20151017_173401.yuv, -1, -1, 60.08891382298475
+Color/0913playSand_Note4_20150913_182713.yuv, -1, -1, 54.20213781634974
+Focus/0911driving_gs5_20150911_170903.yuv, -1, -1, 52.638967951340504
+Artifacts/1019_ColorfulKites2_OppoFind7_VID20150508092524.yuv, -1, -1, 50.97310511796482
+Color/0918_DelMarView_OppoFind7_VID20150409024137.yuv, -1, -1, 60.963433745979756
+Color/0918_DelMarView_GS6_20150921_071851.yuv, -1, -1, 57.249620671155895
+Exposure/0911panGarden_Gs5_20150911_153348.yuv, -1, -1, 30.910356850986815
+Sharpness/0827_OldTownFountain_Note4_20150827_175314.yuv, -1, -1, 66.98799834034504
+Stabilization/0923_FollowPigeons_Note4_20150925_164500.yuv, -1, -1, 52.37350976713892
+Exposure/0911_DogBall_GS6_20150911_185459.yuv, -1, -1, 65.10542900249484
+Focus/1019_HelicopterInSky_OppoFind7_VID20150508093547.yuv, -1, -1, 53.701231222442544
+Artifacts/1006_DucksAndFish2_OppoFind7_VID20150424113906.yuv, -1, -1, 21.403852037578215
+Exposure/1022_QCOMSoccer2_GS6_20151022_174002.yuv, -1, -1, 44.361660121662815
+Color/0912_Cheerleader1_GS5_20150912_153429.yuv, -1, -1, 51.08416769793256
+Stabilization/0923_WalkingToLunch2_HTCOneVx_VIDEO0065.yuv, -1, -1, 22.111602493222765
+Sharpness/0918_GuysPlayVolleyball_LGG2_CAM00804.yuv, -1, -1, 59.486900886871
+Artifacts/1006_FollowTheTrain_LGG2_CAM00834.yuv, -1, -1, 33.939419249644835
+Color/1024_BadmintonAtBalboa_OppoFined7_VID20150512060558.yuv, -1, -1, 56.55278041432343
+Color/0923_RiverBed_Nokia1020_WP_20130305_20_19_29_Pro.yuv, -1, -1, 66.39664147071237
+Stabilization/1019_BollywoodDanceTraining_HTCOneVX_VIDEO0103.yuv, -1, -1, 46.903863263746985
+Focus/1019_SkateBoardingSkills_HTCOneVX_VIDEO0106.yuv, -1, -1, 31.781376267831334
+Sharpness/0826_TrainPullsAway_LG_G2_CAM00726.yuv, -1, -1, 49.77539197067738
+Artifacts/0918_AxelFetchesBall_OppoFind7_VID20150407042625.yuv, -1, -1, 56.47030298915485
+Stabilization/1019_ConstructionBuggy2_OppoFind7_VID20150508085425.yuv, -1, -1, 43.721940313265264
+Artifacts/0918_DogsOnBeach_LGG2_CAM00801.yuv, -1, -1, 60.0145230621533
+Artifacts/0923_ChattingOverLunch_Oppo_VID20150411074312.yuv, -1, -1, 52.436942132562734
+Artifacts/0723_ManUnderTree_Nokia1020_03_WP_20130426_22_12_02_Pro.yuv, -1, -1, 45.67874272310085
+Exposure/0817_WalkIntoBarn_Note4_20150817_110508.yuv, -1, -1, 58.079428695078406
+Sharpness/0911duckSwim_lumia1020_WindowsPhone_20130331_00_06_41_Pro.yuv, -1, -1, 71.3530679089854
+Exposure/1022_QCOMSoccer2_OppoFind7_VID20150510130154.yuv, -1, -1, 45.88425102889353
+Artifacts/1006_PlaneLandsAfar2_GS6_20151006_170428.yuv, -1, -1, 54.47172841710812
+Focus/1019_HelicopterInSky_LGG2_CAM00880.yuv, -1, -1, 49.06343607078643
+Focus/0911driving_note4_20150911_165933.yuv, -1, -1, 66.5912517517975
+Exposure/0911skateBoard_note4_20150911_173356.yuv, -1, -1, 61.56026393111605
+Focus/0913boyWalkBy_Note4_20150913_181950.yuv, -1, -1, 53.97028926187148
+Artifacts/1006_PlaneLandsAfar2_HTCOneVX_VIDEO0099.yuv, -1, -1, 43.13950764717603
+Artifacts/0911nightScene_5s_IMG_0160.yuv, -1, -1, 35.72522155502472
+Stabilization/0923_WalkingToLunch2_Oppo_VID20150411073225.yuv, -1, -1, 34.807565694465076
+Stabilization/0923_SwingingArm_iphone5_IMG_0266.yuv, -1, -1, 59.04852227983286
+Focus/1024_TableTennisAtBalboa_GS6_20151024_104121.yuv, -1, -1, 48.88189507385214
+Artifacts/1019_ColorfulKites2_HTCOneVX_VIDEO0124.yuv, -1, -1, 45.61583176940698
+Sharpness/0923_VerticalAndRollerCoaster_Nokia1020_WP_20130319_18_36_46_Pro.yuv, -1, -1, 63.52586365094227
+Sharpness/0918_GuysPlayVolleyball_OppoFind7_VID20150407045402.yuv, -1, -1, 51.96402419847002
+Artifacts/0918_DogsOnBeach_OppoFind7_VID20150407042141.yuv, -1, -1, 61.24971954276459
+Artifacts/1006_DucksAndFish2_LGG2_CAM00841.yuv, -1, -1, 61.63231290029228
+Exposure/1019_ViewFromMetro2_HTCOneVX_VIDEO0122.yuv, -1, -1, 31.44885665910503
+Sharpness/1006_ParkBus_LGG2_CAM00848.yuv, -1, -1, 49.148682603066575
+Color/1019_FlyingKite3_OppoFind7_VID20150508093248.yuv, -1, -1, 56.46153730303658
+Stabilization/1019_SkateBoarderStepsOut_HTCOneVX.yuv, -1, -1, 36.92437220840707
+Exposure/1019_TrafficFromAfar_LGG2_CAM00850.yuv, -1, -1, 43.920755609079976
+Stabilization/0923_SwingingArm_GS5_20150925_154945.yuv, -1, -1, 59.37209120186363
+Stabilization/1019_SkateBoarderStepsOut_GS6_20151019_182712.yuv, -1, -1, 61.247990668124054
+Focus/1019_HelicopterInSky_HTCOneVX_VIDEO0132.yuv, -1, -1, 16.562095145958942
+Stabilization/1019_BollywoodDanceTraining_LGG2_CAM00851.yuv, -1, -1, 54.376932661964815
+Exposure/0911skateBoard_gs5_20150911_174325.yuv, -1, -1, 49.66995169846537
+Focus/1024_TableTennisAtBalboa_LGG2_CAM00887.yuv, -1, -1, 47.09583764080335
+Sharpness/0923_DucksInWater_Note4_20150925_162109.yuv, -1, -1, 40.52447294642663
+Artifacts/1006_FollowTheTrain_HTCOneVX_VIDEO0086.yuv, -1, -1, 27.486434026248876
+Artifacts/0923_ChattingOverLunch_HTCOneVx_VIDEO0066.yuv, -1, -1, 31.49041491004223
+Sharpness/0913taiZi_Note4_20150913_182629.yuv, -1, -1, 32.936571931353
+Sharpness/0827_OldTownFountain_GS6_20150827_180303.yuv, -1, -1, 66.09356090065711
+Exposure/0911skateBoard_1020_WindowsPhone_20130305_19_55_21_Pro.yuv, -1, -1, 55.13130426481119
+Artifacts/1006_PlaneLandsAfar2_LGG2_CAM00847.yuv, -1, -1, 43.7872800043708
+Sharpness/0923_DucksInWater_GS5_20150925_163115.yuv, -1, -1, 50.962588180142056
+Sharpness/0827_OldTownFountain_iphone5s_IMG_0461.yuv, -1, -1, 63.719020683993385
+Focus/1019_HelicopterInSky_GS6_20151020_141352.yuv, -1, -1, 22.843160682622553
+Exposure/0911skateBoard_5s_IMG_0151.yuv, -1, -1, 66.17622463832596
+Color/0923_RiverBed_Note4_20150925_164759.yuv, -1, -1, 64.23939021565612
+Focus/0723_OldTownShop_GS5_12_20150723_133424.yuv, -1, -1, 51.65560419378202
+Sharpness/0923_DucksInWater_Nokia1020_WP_20130305_19_54_00_Pro.yuv, -1, -1, 54.16160278100095
+Sharpness/0923_DucksInWater_iphone5_IMG_0272.yuv, -1, -1, 49.06737404472014
+Focus/1006_ComplexTrain_LGG2_CAM00835.yuv, -1, -1, 27.479486381827094
+Stabilization/1019_MetroArrives_HTCOneVX_VIDEO0120.yuv, -1, -1, 45.266375356943335

examplar_data_labels/LIVE_Qualcomm/mp4labels.txt

@@ -0,0 +1,208 @@
+Focus/1024_TableTennisAtBalboa_HTCOneVx_VIDEO0138.mp4, -1, -1, 41.64888182462034
+Focus/1019_SkateBoardingSkills_LGG2_CAM00855.mp4, -1, -1, 54.094978137969605
+Focus/1019_SkateBoardingSkills_OppoFind7_VID20150507134738.mp4, -1, -1, 51.02291495174518
+Exposure/0911panGarden_Note4_20150911_152420.mp4, -1, -1, 44.69628201977274
+Artifacts/0923_ChattingOverLunch_G2_CAM00813.mp4, -1, -1, 45.67855937078179
+Artifacts/0911nightScene_note4_20150911_193809.mp4, -1, -1, 36.546723521311996
+Exposure/0912_Burgers_Nokia1020_20130306_17_07_38_Pro.mp4, -1, -1, 64.9396364629356
+Color/1006_EmergencyScene_GS6_20151006_160038.mp4, -1, -1, 58.33545178737855
+Stabilization/0923_WalkingToLunch2_G2_CAM00812.mp4, -1, -1, 48.7769838458931
+Focus/1006_ComplexTrain_HTCOneVX_VIDEO0087.mp4, -1, -1, 22.46553753382192
+Color/0913playSand_GS5_20150913_183702.mp4, -1, -1, 58.51617627453426
+Stabilization/1019_ConstructionBuggy2_GS6_20151020_133229.mp4, -1, -1, 61.262155162968625
+Sharpness/0913taiZi_iPhone5s_IMG_0250.mp4, -1, -1, 29.54546786777324
+Artifacts/1006_FollowTheTrain_OppoFind7_VID20150424111801.mp4, -1, -1, 48.7312983952301
+Stabilization/1019_ConstructionBuggy2_LGG2_CAM00867.mp4, -1, -1, 42.73349382920229
+Artifacts/0918_AxelFetchesBall_GS6_20150919_090332.mp4, -1, -1, 58.23142265391711
+Exposure/1019_ViewFromMetro2_OppoFind7_VID20150508085916.mp4, -1, -1, 36.18976114116138
+Exposure/0923_FollowTheBee_HTCOneVx_VIDEO0072.mp4, -1, -1, 33.04439956164501
+Artifacts/1019_ColorfulKites2_LGG2_CAM00872.mp4, -1, -1, 63.17108813219617
+Exposure/0923_FollowTheBee_GS6_20150923_131355.mp4, -1, -1, 51.45085729576881
+Color/1024_BadmintonAtBalboa_GS6_20151024_104414.mp4, -1, -1, 56.589364068655456
+Exposure/1022_QCOMSoccer2_HTCOneVx_VIDEO0137.mp4, -1, -1, 29.11684686808589
+Sharpness/0912_football_Nokia1020_20130306_17_51_35_Pro.mp4, -1, -1, 57.57359450935232
+Artifacts/0723_ManUnderTree_GS6_03_20150723_130100.mp4, -1, -1, 57.28880925319935
+Exposure/1019_ViewFromMetro2_GS6_20151020_133722.mp4, -1, -1, 39.88244836115969
+Exposure/1019_TrafficFromAfar_OppoFind7_VID20150505125608.mp4, -1, -1, 43.74671562206279
+Artifacts/0918_AxelFetchesBall_HTCOneVX_VIDEO0056.mp4, -1, -1, 42.628796648669706
+Stabilization/0923_ChasingDucksInWater_iphone5_IMG_0274.mp4, -1, -1, 51.11518283501527
+Focus/1006_ComplexTrain_GS6_20151006_155711.mp4, -1, -1, 44.35577617160543
+Artifacts/1006_DucksAndFish2_HTCOneVX_VIDEO0093.mp4, -1, -1, 31.435259054533738
+Color/1024_BadmintonAtBalboa_HTCOneVx_VIDEO0139.mp4, -1, -1, 38.4966864084717
+Artifacts/1006_FollowTheTrain_GS6_20151006_155539.mp4, -1, -1, 47.30872740564676
+Sharpness/0912_football_GS5_20150912_153941.mp4, -1, -1, 53.23486148188893
+Exposure/0911_DogBall_OppoFind7_VID20150330141759.mp4, -1, -1, 66.94118272054322
+Focus/0923_BirdsFlyOnPier_iphone5_IMG_0277.mp4, -1, -1, 24.644612569329546
+Stabilization/1019_MetroArrives_OppoFind7_VID20150508085607.mp4, -1, -1, 55.86662414448337
+Exposure/0911panGarden_Lumia1020_WindowsPhone_20130330_23_52_26_Pro.mp4, -1, -1, 61.25422142212514
+Sharpness/1006_ParkBus_OppoFind7_VID20150424122747.mp4, -1, -1, 55.317108648036935
+Stabilization/1019_ConstructionBuggy2_HTCOneVX_VIDEO0119.mp4, -1, -1, 37.69768941262607
+Sharpness/0918_GuysPlayVolleyball_GS6_20150919_093108.mp4, -1, -1, 65.25654758810917
+Sharpness/1006_ParkBus_GS6_20151006_170532.mp4, -1, -1, 61.532520666710234
+Artifacts/0918_AxelFetchesBall_LGG2_CAM00802.mp4, -1, -1, 52.891660679374006
+Focus/1019_FlyingKite_LGG2_CAM00875.mp4, -1, -1, 42.076217605884914
+Color/1006_FamilyScooter_HTCOneVX_VIDEO0101.mp4, -1, -1, 38.8106247696991
+Color/1006_EmergencyScene_OppoFind7_VID20150424112300.mp4, -1, -1, 62.01071121403491
+Artifacts/0911nightScene_gs5_20150911_194737.mp4, -1, -1, 32.27546017222677
+Exposure/0911panGarden_iPhone5s_IMG_0132.mp4, -1, -1, 56.68920900925609
+Sharpness/0911duckSwim_iphone5s_IMG_0136.mp4, -1, -1, 69.15445557891265
+Stabilization/0923_SwingingArm_Note4_20150925_153938.mp4, -1, -1, 48.1232752921398
+Sharpness/0913taiZi_GS5_20150913_183616.mp4, -1, -1, 32.896692696712435
+Exposure/0911_DogBall_LGG2_CAM00791.mp4, -1, -1, 46.17375037863661
+Exposure/1022_QCOMSoccer2_LGG2_CAM00886.mp4, -1, -1, 26.395621305018185
+Focus/0913boyWalkBy_GS5_20150913_182938.mp4, -1, -1, 51.591805400221055
+Sharpness/0918_GuysPlayVolleyball_HTCOneVX_VIDEO0058.mp4, -1, -1, 53.13309561506308
+Sharpness/0923_VerticalAndRollerCoaster_iphone5_IMG_0258.mp4, -1, -1, 48.66419200608179
+Stabilization/1019_BollywoodDanceTraining_GS6_20151017_175412.mp4, -1, -1, 53.770511195731956
+Exposure/0817_WalkIntoBarn_iPhone5_IMG_0410.mp4, -1, -1, 39.516765825233485
+Color/1006_EmergencyScene_HTCOneVX_VIDEO0090.mp4, -1, -1, 42.55847413392969
+Artifacts/0923_ChattingOverLunch_GS6_20150923_122025.mp4, -1, -1, 60.43118323177181
+Color/0918_CyclistsNear1_OppoFind7_VID20150407033615.mp4, -1, -1, 50.22621242945224
+Stabilization/1019_SkateBoarderStepsOut_OppoFind7_VID20150507134909.mp4, -1, -1, 40.40425472582464
+Stabilization/0923_FollowPigeons_iphone5_IMG_0279.mp4, -1, -1, 47.7605951367109
+Sharpness/0826_TrainPullsAway_OppoFind7_VID20150727213123.mp4, -1, -1, 58.976466922446484
+Artifacts/0723_ManUnderTree_GS5_03_20150723_130016.mp4, -1, -1, 52.56657483287041
+Artifacts/0918_DogsOnBeach_HTCOneVX_VIDEO0055.mp4, -1, -1, 50.378497196338884
+Artifacts/0723_ManUnderTree_HTCOneVx_03_VIDEO0006.mp4, -1, -1, 36.709288593516966
+Color/0912_CheerLeader1_Nokia1080_20130306_17_46_21_Pro.mp4, -1, -1, 41.529339991183875
+Sharpness/0923_VerticalAndRollerCoaster_Note4_20150925_151648.mp4, -1, -1, 55.628924969845
+Color/0923_RiverBed_GS5_20150925_165808.mp4, -1, -1, 61.69082350125231
+Focus/1024_TableTennisAtBalboa_OppoFind7_VID20150512060301.mp4, -1, -1, 50.92808485111922
+Stabilization/0923_FollowPigeons_Nokia1020_WP_20130305_20_16_29_Pro.mp4, -1, -1, 52.181316514970064
+Color/1019_FlyingKite3_LGG2_CAM00877.mp4, -1, -1, 55.74782132350435
+Color/0918_CyclistsNear1_HTCOneVX_VIDEO0050.mp4, -1, -1, 43.90725048088645
+Color/1006_FamilyScooter_OppoFind7_VID20150424122855.mp4, -1, -1, 51.17307547834518
+Artifacts/0911nightScene_1020_WindowsPhone_20130305_21_59_44_Pro.mp4, -1, -1, 21.187371167047402
+Sharpness/0826_TrainPullsAway_iPhone5s_IMG_0423.mp4, -1, -1, 60.880133455554756
+Stabilization/0923_CrazySubmarine_Note4_20150925_160005.mp4, -1, -1, 56.08844663824663
+Focus/0923_BirdsFlyOnPier_Note4_20150925_164318.mp4, -1, -1, 34.77241986411562
+Artifacts/1006_DucksAndFish2_GS6_20151006_161642.mp4, -1, -1, 51.0393821878597
+Stabilization/1019_SkateBoarderStepsOut_LGG2_CAM00857.mp4, -1, -1, 48.672533412525446
+Exposure/0911_DogBall_HTCOneVx_VIDEO0045.mp4, -1, -1, 49.7123541769036
+Color/1019_FlyingKite3_HTCOneVX_VIDEO0129.mp4, -1, -1, 49.33731661081409
+Focus/1006_ComplexTrain_OppoFind7_VID20150424111934.mp4, -1, -1, 31.508494371587865
+Focus/0723_OldTownShop_GS6_12_20150723_133507.mp4, -1, -1, 64.9917581345629
+Stabilization/0923_ChasingDucksInWater_GS5_20150925_163521.mp4, -1, -1, 48.91083829259606
+Focus/0723_OldTownShop_Nokia1020_12_WP_20130426_22_46_05_Pro.mp4, -1, -1, 52.969252792370625
+Focus/1019_FlyingKite_HTCOneVX_VIDEO0127.mp4, -1, -1, 20.802340977962967
+Color/1006_FamilyScooter_LGG2_CAM00849.mp4, -1, -1, 49.68892718051139
+Stabilization/0923_ChasingDucksInWater_Nokia1020_WP_20130305_19_56_39_Pro.mp4, -1, -1, 40.700157633712266
+Sharpness/0827_OldTownFountain_GS5_20150827_180240.mp4, -1, -1, 66.22997014329692
+Exposure/0912_Burgers_GS5_20150912_145544.mp4, -1, -1, 63.475870057217286
+Color/1006_FamilyScooter_GS6_20151006_170632.mp4, -1, -1, 60.28464598447542
+Color/1024_BadmintonAtBalboa_LGG2_CAM00888.mp4, -1, -1, 52.3694844910836
+Stabilization/1019_BollywoodDanceTraining_OppoFind7_VID20150505131616.mp4, -1, -1, 49.18933619585497
+Exposure/0912_Burgers_Note4_20150912_144558.mp4, -1, -1, 61.98320494013754
+Stabilization/0923_WalkingToLunch2_GS6_20150923_120934.mp4, -1, -1, 46.50669762802673
+Stabilization/1019_MetroArrives_LGG2_CAM00868.mp4, -1, -1, 59.05090948407636
+Artifacts/1019_ColorfulKites2_GS6_20151020_140327.mp4, -1, -1, 69.34239333030179
+Focus/0923_BirdsFlyOnPier_GS5_20150925_165328.mp4, -1, -1, 46.858952126850156
+Focus/0923_BirdsFlyOnPier_Nokia1020_WP_20130305_20_14_47_Pro.mp4, -1, -1, 49.44336625534995
+Color/0923_RiverBed_iphone5_IMG_0280.mp4, -1, -1, 63.16961373461168
+Sharpness/0911duckSwim_gs5_20150911_154802.mp4, -1, -1, 59.46925058712509
+Color/0918_DelMarView_HTCOneVX_VIDEO0062.mp4, -1, -1, 45.79645003864048
+Color/0918_CyclistsNear1_LGG2_CAM00795.mp4, -1, -1, 46.753110388198024
+Exposure/0817_WalkIntoBarn_OppoFind7_VID20150718144333.mp4, -1, -1, 49.02405995062579
+Color/1019_FlyingKite3_GS6_20151020_141052.mp4, -1, -1, 73.64276415152705
+Artifacts/0918_DogsOnBeach_GS6_20150919_085853.mp4, -1, -1, 65.94233767581703
+Stabilization/0923_CrazySubmarine_Nokia1020_WP_20130305_19_31_30_Pro.mp4, -1, -1, 59.038670696087216
+Focus/0913boyWalkBy_iPhone5s_IMG_0245.mp4, -1, -1, 62.812549213807614
+Stabilization/1019_MetroArrives_GS6_20151020_133409.mp4, -1, -1, 65.82583000957678
+Sharpness/0911duckSwim_notes4_20150911_153835.mp4, -1, -1, 71.74875099099955
+Sharpness/1006_ParkBus_HTCOneVX_VIDEO0100.mp4, -1, -1, 31.50387615900128
+Focus/1019_SkateBoardingSkills_GS6_20151019_182541.mp4, -1, -1, 59.007584920405066
+Color/0918_DelMarView_LGG2_CAM00809.mp4, -1, -1, 48.7408776505235
+Color/0912_Cheerleader1_Note4_20150912_152444.mp4, -1, -1, 46.853275685229434
+Focus/1019_FlyingKite_OppoFind7_VID20150508093000.mp4, -1, -1, 54.075938062059876
+Exposure/1019_TrafficFromAfar_HTCOneVX_VIDEO0102.mp4, -1, -1, 38.676262337846296
+Focus/0911driving_5s_IMG_0145.mp4, -1, -1, 44.9586564442459
+Exposure/1019_ViewFromMetro2_LGG2_CAM00870.mp4, -1, -1, 37.58488437787315
+Focus/0723_OldTownShop_HTCOneVx_12_VIDEO0019.mp4, -1, -1, 30.53679255118817
+Stabilization/0923_ChasingDucksInWater_Note4_20150925_162515.mp4, -1, -1, 48.75250168989043
+Sharpness/0826_TrainPullsAway_Note4_20150826_175232.mp4, -1, -1, 57.182744558097575
+Focus/1019_FlyingKite_GS6_20151020_140804.mp4, -1, -1, 23.50289297893303
+Color/0918_CyclistsNear1_GS6_20150919_081325.mp4, -1, -1, 52.81500748332389
+Artifacts/1006_PlaneLandsAfar2_OppoFind7_VID20150424122651.mp4, -1, -1, 57.22288718627244
+Exposure/0912_Burgers_iphone5s_IMG_0167.mp4, -1, -1, 64.38345299460406
+Sharpness/0923_VerticalAndRollerCoaster_GS5_20150925_152655.mp4, -1, -1, 55.74790371519573
+Stabilization/0923_FollowPigeons_GS5_20150925_165510.mp4, -1, -1, 55.46436837100166
+Exposure/0817_WalkIntoBarn_LGG2_CAM00718.mp4, -1, -1, 51.146870112855
+Color/0913playSand_iPhone5s_IMG_0251.mp4, -1, -1, 72.96615929361013
+Stabilization/0923_CrazySubmarine_GS5_20150925_161013.mp4, -1, -1, 62.68393406357753
+Sharpness/0912_football_Note4_20150912_152954.mp4, -1, -1, 55.629277646191774
+Color/1006_EmergencyScene_LGG2_CAM00838.mp4, -1, -1, 66.8679745839632
+Stabilization/0923_CrazySubmarine_iphone5_IMG_0268.mp4, -1, -1, 62.539356287114956
+Color/0912_Cheerleader1_iphone_IMG_0184.mp4, -1, -1, 59.46973486351669
+Exposure/1019_TrafficFromAfar_GS6_20151017_173401.mp4, -1, -1, 60.08891382298475
+Color/0913playSand_Note4_20150913_182713.mp4, -1, -1, 54.20213781634974
+Focus/0911driving_gs5_20150911_170903.mp4, -1, -1, 52.638967951340504
+Artifacts/1019_ColorfulKites2_OppoFind7_VID20150508092524.mp4, -1, -1, 50.97310511796482
+Color/0918_DelMarView_OppoFind7_VID20150409024137.mp4, -1, -1, 60.963433745979756
+Color/0918_DelMarView_GS6_20150921_071851.mp4, -1, -1, 57.249620671155895
+Exposure/0911panGarden_Gs5_20150911_153348.mp4, -1, -1, 30.910356850986815
+Sharpness/0827_OldTownFountain_Note4_20150827_175314.mp4, -1, -1, 66.98799834034504
+Stabilization/0923_FollowPigeons_Note4_20150925_164500.mp4, -1, -1, 52.37350976713892
+Exposure/0911_DogBall_GS6_20150911_185459.mp4, -1, -1, 65.10542900249484
+Focus/1019_HelicopterInSky_OppoFind7_VID20150508093547.mp4, -1, -1, 53.701231222442544
+Artifacts/1006_DucksAndFish2_OppoFind7_VID20150424113906.mp4, -1, -1, 21.403852037578215
+Exposure/1022_QCOMSoccer2_GS6_20151022_174002.mp4, -1, -1, 44.361660121662815
+Color/0912_Cheerleader1_GS5_20150912_153429.mp4, -1, -1, 51.08416769793256
+Stabilization/0923_WalkingToLunch2_HTCOneVx_VIDEO0065.mp4, -1, -1, 22.111602493222765
+Sharpness/0918_GuysPlayVolleyball_LGG2_CAM00804.mp4, -1, -1, 59.486900886871
+Artifacts/1006_FollowTheTrain_LGG2_CAM00834.mp4, -1, -1, 33.939419249644835
+Color/1024_BadmintonAtBalboa_OppoFined7_VID20150512060558.mp4, -1, -1, 56.55278041432343
+Color/0923_RiverBed_Nokia1020_WP_20130305_20_19_29_Pro.mp4, -1, -1, 66.39664147071237
+Stabilization/1019_BollywoodDanceTraining_HTCOneVX_VIDEO0103.mp4, -1, -1, 46.903863263746985
+Focus/1019_SkateBoardingSkills_HTCOneVX_VIDEO0106.mp4, -1, -1, 31.781376267831334
+Sharpness/0826_TrainPullsAway_LG_G2_CAM00726.mp4, -1, -1, 49.77539197067738
+Artifacts/0918_AxelFetchesBall_OppoFind7_VID20150407042625.mp4, -1, -1, 56.47030298915485
+Stabilization/1019_ConstructionBuggy2_OppoFind7_VID20150508085425.mp4, -1, -1, 43.721940313265264
+Artifacts/0918_DogsOnBeach_LGG2_CAM00801.mp4, -1, -1, 60.0145230621533
+Artifacts/0923_ChattingOverLunch_Oppo_VID20150411074312.mp4, -1, -1, 52.436942132562734
+Artifacts/0723_ManUnderTree_Nokia1020_03_WP_20130426_22_12_02_Pro.mp4, -1, -1, 45.67874272310085
+Exposure/0817_WalkIntoBarn_Note4_20150817_110508.mp4, -1, -1, 58.079428695078406
+Sharpness/0911duckSwim_lumia1020_WindowsPhone_20130331_00_06_41_Pro.mp4, -1, -1, 71.3530679089854
+Exposure/1022_QCOMSoccer2_OppoFind7_VID20150510130154.mp4, -1, -1, 45.88425102889353
+Artifacts/1006_PlaneLandsAfar2_GS6_20151006_170428.mp4, -1, -1, 54.47172841710812
+Focus/1019_HelicopterInSky_LGG2_CAM00880.mp4, -1, -1, 49.06343607078643
+Focus/0911driving_note4_20150911_165933.mp4, -1, -1, 66.5912517517975
+Exposure/0911skateBoard_note4_20150911_173356.mp4, -1, -1, 61.56026393111605
+Focus/0913boyWalkBy_Note4_20150913_181950.mp4, -1, -1, 53.97028926187148
+Artifacts/1006_PlaneLandsAfar2_HTCOneVX_VIDEO0099.mp4, -1, -1, 43.13950764717603
+Artifacts/0911nightScene_5s_IMG_0160.mp4, -1, -1, 35.72522155502472
+Stabilization/0923_WalkingToLunch2_Oppo_VID20150411073225.mp4, -1, -1, 34.807565694465076
+Stabilization/0923_SwingingArm_iphone5_IMG_0266.mp4, -1, -1, 59.04852227983286
+Focus/1024_TableTennisAtBalboa_GS6_20151024_104121.mp4, -1, -1, 48.88189507385214
+Artifacts/1019_ColorfulKites2_HTCOneVX_VIDEO0124.mp4, -1, -1, 45.61583176940698
+Sharpness/0923_VerticalAndRollerCoaster_Nokia1020_WP_20130319_18_36_46_Pro.mp4, -1, -1, 63.52586365094227
+Sharpness/0918_GuysPlayVolleyball_OppoFind7_VID20150407045402.mp4, -1, -1, 51.96402419847002
+Artifacts/0918_DogsOnBeach_OppoFind7_VID20150407042141.mp4, -1, -1, 61.24971954276459
+Artifacts/1006_DucksAndFish2_LGG2_CAM00841.mp4, -1, -1, 61.63231290029228
+Exposure/1019_ViewFromMetro2_HTCOneVX_VIDEO0122.mp4, -1, -1, 31.44885665910503
+Sharpness/1006_ParkBus_LGG2_CAM00848.mp4, -1, -1, 49.148682603066575
+Color/1019_FlyingKite3_OppoFind7_VID20150508093248.mp4, -1, -1, 56.46153730303658
+Stabilization/1019_SkateBoarderStepsOut_HTCOneVX.mp4, -1, -1, 36.92437220840707
+Exposure/1019_TrafficFromAfar_LGG2_CAM00850.mp4, -1, -1, 43.920755609079976
+Stabilization/0923_SwingingArm_GS5_20150925_154945.mp4, -1, -1, 59.37209120186363
+Stabilization/1019_SkateBoarderStepsOut_GS6_20151019_182712.mp4, -1, -1, 61.247990668124054
+Focus/1019_HelicopterInSky_HTCOneVX_VIDEO0132.mp4, -1, -1, 16.562095145958942
+Stabilization/1019_BollywoodDanceTraining_LGG2_CAM00851.mp4, -1, -1, 54.376932661964815
+Exposure/0911skateBoard_gs5_20150911_174325.mp4, -1, -1, 49.66995169846537
+Focus/1024_TableTennisAtBalboa_LGG2_CAM00887.mp4, -1, -1, 47.09583764080335
+Sharpness/0923_DucksInWater_Note4_20150925_162109.mp4, -1, -1, 40.52447294642663
+Artifacts/1006_FollowTheTrain_HTCOneVX_VIDEO0086.mp4, -1, -1, 27.486434026248876
+Artifacts/0923_ChattingOverLunch_HTCOneVx_VIDEO0066.mp4, -1, -1, 31.49041491004223
+Sharpness/0913taiZi_Note4_20150913_182629.mp4, -1, -1, 32.936571931353
+Sharpness/0827_OldTownFountain_GS6_20150827_180303.mp4, -1, -1, 66.09356090065711
+Exposure/0911skateBoard_1020_WindowsPhone_20130305_19_55_21_Pro.mp4, -1, -1, 55.13130426481119
+Artifacts/1006_PlaneLandsAfar2_LGG2_CAM00847.mp4, -1, -1, 43.7872800043708
+Sharpness/0923_DucksInWater_GS5_20150925_163115.mp4, -1, -1, 50.962588180142056
+Sharpness/0827_OldTownFountain_iphone5s_IMG_0461.mp4, -1, -1, 63.719020683993385
+Focus/1019_HelicopterInSky_GS6_20151020_141352.mp4, -1, -1, 22.843160682622553
+Exposure/0911skateBoard_5s_IMG_0151.mp4, -1, -1, 66.17622463832596
+Color/0923_RiverBed_Note4_20150925_164759.mp4, -1, -1, 64.23939021565612
+Focus/0723_OldTownShop_GS5_12_20150723_133424.mp4, -1, -1, 51.65560419378202
+Sharpness/0923_DucksInWater_Nokia1020_WP_20130305_19_54_00_Pro.mp4, -1, -1, 54.16160278100095
+Sharpness/0923_DucksInWater_iphone5_IMG_0272.mp4, -1, -1, 49.06737404472014
+Focus/1006_ComplexTrain_LGG2_CAM00835.mp4, -1, -1, 27.479486381827094
+Stabilization/1019_MetroArrives_HTCOneVX_VIDEO0120.mp4, -1, -1, 45.266375356943335

examplar_data_labels/LIVE_VQA/labels.txt

@@ -0,0 +1,148 @@
+pa/pa2_25fps.yuv, -1, -1, 44.5104
+pa/pa3_25fps.yuv, -1, -1, 70.1054
+pa/pa4_25fps.yuv, -1, -1, 66.4280
+pa/pa5_25fps.yuv, -1, -1, 75.1225
+pa/pa6_25fps.yuv, -1, -1, 73.8803
+pa/pa7_25fps.yuv, -1, -1, 63.2564
+pa/pa8_25fps.yuv, -1, -1, 61.2726
+pa/pa9_25fps.yuv, -1, -1, 40.5551
+pa/pa10_25fps.yuv, -1, -1, 52.6111
+pa/pa11_25fps.yuv, -1, -1, 60.2534
+pa/pa12_25fps.yuv, -1, -1, 68.7186
+pa/pa13_25fps.yuv, -1, -1, 42.9784
+pa/pa14_25fps.yuv, -1, -1, 51.0530
+pa/pa15_25fps.yuv, -1, -1, 55.7020
+pa/pa16_25fps.yuv, -1, -1, 65.6457
+rb/rb2_25fps.yuv, -1, -1, 64.9369
+rb/rb3_25fps.yuv, -1, -1, 46.2446
+rb/rb4_25fps.yuv, -1, -1, 54.3732
+rb/rb5_25fps.yuv, -1, -1, 46.4907
+rb/rb6_25fps.yuv, -1, -1, 68.1064
+rb/rb7_25fps.yuv, -1, -1, 54.8101
+rb/rb8_25fps.yuv, -1, -1, 54.6555
+rb/rb9_25fps.yuv, -1, -1, 39.1978
+rb/rb10_25fps.yuv, -1, -1, 43.6833
+rb/rb11_25fps.yuv, -1, -1, 55.8563
+rb/rb12_25fps.yuv, -1, -1, 63.5809
+rb/rb13_25fps.yuv, -1, -1, 38.8828
+rb/rb14_25fps.yuv, -1, -1, 45.6069
+rb/rb15_25fps.yuv, -1, -1, 48.0089
+rb/rb16_25fps.yuv, -1, -1, 47.5270
+rh/rh2_25fps.yuv, -1, -1, 68.1431
+rh/rh3_25fps.yuv, -1, -1, 63.5698
+rh/rh4_25fps.yuv, -1, -1, 48.0196
+rh/rh5_25fps.yuv, -1, -1, 51.4980
+rh/rh6_25fps.yuv, -1, -1, 55.2291
+rh/rh7_25fps.yuv, -1, -1, 62.3778
+rh/rh8_25fps.yuv, -1, -1, 42.6909
+rh/rh9_25fps.yuv, -1, -1, 37.8713
+rh/rh10_25fps.yuv, -1, -1, 45.4363
+rh/rh11_25fps.yuv, -1, -1, 53.6343
+rh/rh14_25fps.yuv, -1, -1, 42.8568
+rh/rh15_25fps.yuv, -1, -1, 52.0988
+rh/rh16_25fps.yuv, -1, -1, 62.2062
+tr/tr2_25fps.yuv, -1, -1, 71.2731
+tr/tr3_25fps.yuv, -1, -1, 72.1356
+tr/tr4_25fps.yuv, -1, -1, 64.6561
+tr/tr5_25fps.yuv, -1, -1, 53.1125
+tr/tr6_25fps.yuv, -1, -1, 73.4730
+tr/tr7_25fps.yuv, -1, -1, 55.3531
+tr/tr8_25fps.yuv, -1, -1, 52.4524
+tr/tr9_25fps.yuv, -1, -1, 38.6726
+tr/tr10_25fps.yuv, -1, -1, 47.7716
+tr/tr11_25fps.yuv, -1, -1, 56.9119
+tr/tr12_25fps.yuv, -1, -1, 63.7984
+tr/tr13_25fps.yuv, -1, -1, 33.4734
+tr/tr14_25fps.yuv, -1, -1, 42.5381
+tr/tr15_25fps.yuv, -1, -1, 56.1328
+tr/tr16_25fps.yuv, -1, -1, 65.7102
+st/st2_25fps.yuv, -1, -1, 65.6522
+st/st3_25fps.yuv, -1, -1, 61.3221
+st/st4_25fps.yuv, -1, -1, 44.0305
+st/st5_25fps.yuv, -1, -1, 41.4157
+st/st6_25fps.yuv, -1, -1, 58.4534
+st/st7_25fps.yuv, -1, -1, 44.2762
+st/st8_25fps.yuv, -1, -1, 48.3834
+st/st9_25fps.yuv, -1, -1, 40.7745
+st/st10_25fps.yuv, -1, -1, 46.5633
+st/st11_25fps.yuv, -1, -1, 52.3269
+st/st12_25fps.yuv, -1, -1, 56.0811
+st/st13_25fps.yuv, -1, -1, 36.5136
+st/st14_25fps.yuv, -1, -1, 42.9632
+st/st15_25fps.yuv, -1, -1, 49.1987
+st/st16_25fps.yuv, -1, -1, 57.4200
+sf/sf2_25fps.yuv, -1, -1, 54.9213
+sf/sf3_25fps.yuv, -1, -1, 63.2756
+sf/sf4_25fps.yuv, -1, -1, 56.8614
+sf/sf5_25fps.yuv, -1, -1, 49.2987
+sf/sf6_25fps.yuv, -1, -1, 59.3959
+sf/sf7_25fps.yuv, -1, -1, 44.8094
+sf/sf8_25fps.yuv, -1, -1, 39.1088
+sf/sf9_25fps.yuv, -1, -1, 32.6002
+sf/sf10_25fps.yuv, -1, -1, 44.0164
+sf/sf11_25fps.yuv, -1, -1, 54.9423
+sf/sf12_25fps.yuv, -1, -1, 57.1497
+sf/sf13_25fps.yuv, -1, -1, 40.9999
+sf/sf14_25fps.yuv, -1, -1, 44.6477
+sf/sf15_25fps.yuv, -1, -1, 49.2215
+sf/sf16_25fps.yuv, -1, -1, 53.7003
+bs/bs2_25fps.yuv, -1, -1, 68.9412
+bs/bs3_25fps.yuv, -1, -1, 52.9363
+bs/bs4_25fps.yuv, -1, -1, 51.0109
+bs/bs5_25fps.yuv, -1, -1, 55.9066
+bs/bs6_25fps.yuv, -1, -1, 61.7965
+bs/bs7_25fps.yuv, -1, -1, 45.9273
+bs/bs8_25fps.yuv, -1, -1, 40.9576
+bs/bs9_25fps.yuv, -1, -1, 31.9421
+bs/bs10_25fps.yuv, -1, -1, 36.6396
+bs/bs11_25fps.yuv, -1, -1, 38.6448
+bs/bs12_25fps.yuv, -1, -1, 52.1844
+bs/bs13_25fps.yuv, -1, -1, 32.7252
+bs/bs14_25fps.yuv, -1, -1, 43.9984
+bs/bs15_25fps.yuv, -1, -1, 50.5090
+bs/bs16_25fps.yuv, -1, -1, 53.4364
+sh/sh2_50fps.yuv, -1, -1, 81.1601
+sh/sh3_50fps.yuv, -1, -1, 70.5494
+sh/sh4_50fps.yuv, -1, -1, 54.9174
+sh/sh5_50fps.yuv, -1, -1, 49.6350
+sh/sh6_50fps.yuv, -1, -1, 55.5307
+sh/sh7_50fps.yuv, -1, -1, 61.2837
+sh/sh8_50fps.yuv, -1, -1, 46.2254
+sh/sh9_50fps.yuv, -1, -1, 36.2440
+sh/sh10_50fps.yuv, -1, -1, 40.8004
+sh/sh11_50fps.yuv, -1, -1, 51.6153
+sh/sh12_50fps.yuv, -1, -1, 66.3166
+sh/sh13_50fps.yuv, -1, -1, 37.0212
+sh/sh14_50fps.yuv, -1, -1, 44.0813
+sh/sh15_50fps.yuv, -1, -1, 57.5757
+sh/sh16_50fps.yuv, -1, -1, 62.0745
+mc/mc2_50fps.yuv, -1, -1, 78.3431
+mc/mc3_50fps.yuv, -1, -1, 69.2258
+mc/mc4_50fps.yuv, -1, -1, 59.5299
+mc/mc5_50fps.yuv, -1, -1, 57.8482
+mc/mc6_50fps.yuv, -1, -1, 73.3075
+mc/mc7_50fps.yuv, -1, -1, 58.5392
+mc/mc8_50fps.yuv, -1, -1, 54.0963
+mc/mc9_50fps.yuv, -1, -1, 47.3711
+mc/mc10_50fps.yuv, -1, -1, 48.7705
+mc/mc11_50fps.yuv, -1, -1, 57.6788
+mc/mc12_50fps.yuv, -1, -1, 67.8232
+mc/mc13_50fps.yuv, -1, -1, 30.9426
+mc/mc14_50fps.yuv, -1, -1, 40.5326
+mc/mc15_50fps.yuv, -1, -1, 52.5435
+mc/mc16_50fps.yuv, -1, -1, 64.8173
+pr/pr2_50fps.yuv, -1, -1, 61.3882
+pr/pr3_50fps.yuv, -1, -1, 66.3322
+pr/pr4_50fps.yuv, -1, -1, 45.4702
+pr/pr5_50fps.yuv, -1, -1, 45.3150
+pr/pr6_50fps.yuv, -1, -1, 55.3240
+pr/pr7_50fps.yuv, -1, -1, 56.1730
+pr/pr8_50fps.yuv, -1, -1, 44.6086
+pr/pr9_50fps.yuv, -1, -1, 39.8067
+pr/pr10_50fps.yuv, -1, -1, 53.7598
+pr/pr11_50fps.yuv, -1, -1, 59.8921
+pr/pr12_50fps.yuv, -1, -1, 77.2518
+pr/pr13_50fps.yuv, -1, -1, 39.7105
+pr/pr14_50fps.yuv, -1, -1, 46.8271
+pr/pr15_50fps.yuv, -1, -1, 54.4239
+pr/pr16_50fps.yuv, -1, -1, 61.8235

examplar_data_labels/LIVE_VQA/names.txt

@@ -0,0 +1,150 @@
+pa2_25fps.yuv
+pa3_25fps.yuv
+pa4_25fps.yuv
+pa5_25fps.yuv
+pa6_25fps.yuv
+pa7_25fps.yuv
+pa8_25fps.yuv
+pa9_25fps.yuv
+pa10_25fps.yuv
+pa11_25fps.yuv
+pa12_25fps.yuv
+pa13_25fps.yuv
+pa14_25fps.yuv
+pa15_25fps.yuv
+pa16_25fps.yuv
+rb2_25fps.yuv
+rb3_25fps.yuv
+rb4_25fps.yuv
+rb5_25fps.yuv
+rb6_25fps.yuv
+rb7_25fps.yuv
+rb8_25fps.yuv
+rb9_25fps.yuv
+rb10_25fps.yuv
+rb11_25fps.yuv
+rb12_25fps.yuv
+rb13_25fps.yuv
+rb14_25fps.yuv
+rb15_25fps.yuv
+rb16_25fps.yuv
+rh2_25fps.yuv
+rh3_25fps.yuv
+rh4_25fps.yuv
+rh5_25fps.yuv
+rh6_25fps.yuv
+rh7_25fps.yuv
+rh8_25fps.yuv
+rh9_25fps.yuv
+rh10_25fps.yuv
+rh11_25fps.yuv
+rh12_25fps.yuv
+rh13_25fps.yuv
+rh14_25fps.yuv
+rh15_25fps.yuv
+rh16_25fps.yuv
+tr2_25fps.yuv
+tr3_25fps.yuv
+tr4_25fps.yuv
+tr5_25fps.yuv
+tr6_25fps.yuv
+tr7_25fps.yuv
+tr8_25fps.yuv
+tr9_25fps.yuv
+tr10_25fps.yuv
+tr11_25fps.yuv
+tr12_25fps.yuv
+tr13_25fps.yuv
+tr14_25fps.yuv
+tr15_25fps.yuv
+tr16_25fps.yuv
+st2_25fps.yuv
+st3_25fps.yuv
+st4_25fps.yuv
+st5_25fps.yuv
+st6_25fps.yuv
+st7_25fps.yuv
+st8_25fps.yuv
+st9_25fps.yuv
+st10_25fps.yuv
+st11_25fps.yuv
+st12_25fps.yuv
+st13_25fps.yuv
+st14_25fps.yuv
+st15_25fps.yuv
+st16_25fps.yuv
+sf2_25fps.yuv
+sf3_25fps.yuv
+sf4_25fps.yuv
+sf5_25fps.yuv
+sf6_25fps.yuv
+sf7_25fps.yuv
+sf8_25fps.yuv
+sf9_25fps.yuv
+sf10_25fps.yuv
+sf11_25fps.yuv
+sf12_25fps.yuv
+sf13_25fps.yuv
+sf14_25fps.yuv
+sf15_25fps.yuv
+sf16_25fps.yuv
+bs2_25fps.yuv
+bs3_25fps.yuv
+bs4_25fps.yuv
+bs5_25fps.yuv
+bs6_25fps.yuv
+bs7_25fps.yuv
+bs8_25fps.yuv
+bs9_25fps.yuv
+bs10_25fps.yuv
+bs11_25fps.yuv
+bs12_25fps.yuv
+bs13_25fps.yuv
+bs14_25fps.yuv
+bs15_25fps.yuv
+bs16_25fps.yuv
+sh2_50fps.yuv
+sh3_50fps.yuv
+sh4_50fps.yuv
+sh5_50fps.yuv
+sh6_50fps.yuv
+sh7_50fps.yuv
+sh8_50fps.yuv
+sh9_50fps.yuv
+sh10_50fps.yuv
+sh11_50fps.yuv
+sh12_50fps.yuv
+sh13_50fps.yuv
+sh14_50fps.yuv
+sh15_50fps.yuv
+sh16_50fps.yuv
+mc2_50fps.yuv
+mc3_50fps.yuv
+mc4_50fps.yuv
+mc5_50fps.yuv
+mc6_50fps.yuv
+mc7_50fps.yuv
+mc8_50fps.yuv
+mc9_50fps.yuv
+mc10_50fps.yuv
+mc11_50fps.yuv
+mc12_50fps.yuv
+mc13_50fps.yuv
+mc14_50fps.yuv
+mc15_50fps.yuv
+mc16_50fps.yuv
+pr2_50fps.yuv
+pr3_50fps.yuv
+pr4_50fps.yuv
+pr5_50fps.yuv
+pr6_50fps.yuv
+pr7_50fps.yuv
+pr8_50fps.yuv
+pr9_50fps.yuv
+pr10_50fps.yuv
+pr11_50fps.yuv
+pr12_50fps.yuv
+pr13_50fps.yuv
+pr14_50fps.yuv
+pr15_50fps.yuv
+pr16_50fps.yuv

examplar_data_labels/LIVE_VQA/scores.txt

@@ -0,0 +1,150 @@
+44.5104	12.2909
+70.1054	8.4630
+66.4280	10.9220
+75.1225	8.7056
+73.8803	5.7825
+63.2564	8.8315
+61.2726	10.6827
+40.5551	8.4040
+52.6111	9.8646
+60.2534	9.0097
+68.7186	9.3995
+42.9784	8.5050
+51.0530	8.0119
+55.7020	9.3731
+65.6457	10.8023
+64.9369	12.4744
+46.2446	9.8897
+54.3732	12.0351
+46.4907	10.9136
+68.1064	10.4983
+54.8101	13.2412
+54.6555	12.2369
+39.1978	11.7595
+43.6833	12.6685
+55.8563	15.2382
+63.5809	12.0636
+38.8828	11.0500
+45.6069	14.4528
+48.0089	13.7996
+47.5270	11.8475
+68.1431	12.0123
+63.5698	12.6835
+48.0196	11.2378
+51.4980	13.1559
+55.2291	11.2665
+62.3778	12.1601
+42.6909	9.5547
+37.8713	9.9518
+45.4363	11.9058
+53.6343	13.7169
+62.9934	10.0094
+31.4716	8.0896
+42.8568	11.4820
+52.0988	8.0925
+62.2062	10.8021
+71.2731	7.3171
+72.1356	8.2769
+64.6561	8.7193
+53.1125	10.2891
+73.4730	11.2189
+55.3531	10.7032
+52.4524	9.9872
+38.6726	8.7816
+47.7716	8.6263
+56.9119	9.3595
+63.7984	7.4827
+33.4734	8.8625
+42.5381	12.2394
+56.1328	10.0524
+65.7102	10.8513
+65.6522	11.8297
+61.3221	11.2218
+44.0305	12.3100
+41.4157	10.1887
+58.4534	10.2342
+44.2762	10.2308
+48.3834	10.8759
+40.7745	10.9440
+46.5633	9.3641
+52.3269	11.2327
+56.0811	9.9024
+36.5136	10.6661
+42.9632	9.5615
+49.1987	12.5682
+57.4200	10.8714
+54.9213	9.9593
+63.2756	7.0135
+56.8614	10.3063
+49.2987	7.9941
+59.3959	8.3076
+44.8094	11.1511
+39.1088	8.8315
+32.6002	7.5710
+44.0164	9.5158
+54.9423	8.8703
+57.1497	10.3586
+40.9999	10.2129
+44.6477	9.6876
+49.2215	8.2303
+53.7003	8.3839
+68.9412	13.2694
+52.9363	10.9429
+51.0109	11.6969
+55.9066	12.9653
+61.7965	8.9395
+45.9273	12.2075
+40.9576	10.0565
+31.9421	10.0953
+36.6396	10.2083
+38.6448	9.1071
+52.1844	10.8366
+32.7252	11.6010
+43.9984	9.6540
+50.5090	8.9686
+53.4364	11.3882
+81.1601	8.8839
+70.5494	7.0154
+54.9174	10.3442
+49.6350	9.8661
+55.5307	8.4316
+61.2837	9.8106
+46.2254	8.0034
+36.2440	8.7969
+40.8004	8.8023
+51.6153	10.4552
+66.3166	10.0913
+37.0212	7.4451
+44.0813	9.4971
+57.5757	6.4381
+62.0745	6.2390
+78.3431	9.9876
+69.2258	8.0969
+59.5299	9.8755
+57.8482	10.2606
+73.3075	9.0790
+58.5392	11.3208
+54.0963	10.0428
+47.3711	10.8012
+48.7705	7.7892
+57.6788	9.7494
+67.8232	7.4454
+30.9426	8.0339
+40.5326	9.8009
+52.5435	9.9240
+64.8173	9.5076
+61.3882	10.2155
+66.3322	11.1123
+45.4702	7.6892
+45.3150	8.6377
+55.3240	6.1770
+56.1730	8.7040
+44.6086	10.3585
+39.8067	8.2885
+53.7598	9.0671
+59.8921	10.6386
+77.2518	8.7931
+39.7105	9.5447
+46.8271	10.3513
+54.4239	11.2077
+61.8235	11.1164

examplar_data_labels/LIVE_VQC/labels.txt

@@ -0,0 +1,585 @@
+Video/A001.mp4, 10.002367, 30.0, 80.232
+Video/A002.mp4, 10.04, 29.97002997002997, 57.3005
+Video/A003.mp4, 10.008333, 29.97002997002997, 78.1558
+Video/A004.mp4, 9.966821999999999, 30.0, 59.3179
+Video/A005.mp4, 10.033332999999999, 30.0, 47.1778
+Video/A006.mp4, 10.030721999999999, 30.0, 84.6517
+Video/A007.mp4, 10.008333, 29.97002997002997, 50.8675
+Video/A008.mp4, 10.006667, 29.97002997002997, 46.8902
+Video/A009.mp4, 10.006667, 29.97002997002997, 80.7925
+Video/A010.mp4, 10.0306, 30.0, 31.2785
+Video/A011.mp4, 10.006611, 30.0, 87.2287
+Video/A012.mp4, 10.033389, 29.916666666666668, 47.7826
+Video/A013.mp4, 10.04, 29.97002997002997, 83.4083
+Video/A014.mp4, 10.006667, 29.97002997002997, 69.9652
+Video/A015.mp4, 10.003910999999999, 30.0, 78.3957
+Video/A016.mp4, 10.000855999999999, 29.916666666666668, 37.8421
+Video/A017.mp4, 10.030332999999999, 30.0, 65.4778
+Video/A018.mp4, 10.006667, 29.97002997002997, 76.0205
+Video/A019.mp4, 10.04, 29.97002997002997, 70.843
+Video/A020.mp4, 10.005455999999999, 30.0, 77.7258
+Video/A021.mp4, 10.016733, 29.850746268656717, 37.6011
+Video/A022.mp4, 10.031666999999999, 30.0, 38.674
+Video/A023.mp4, 10.006667, 29.97002997002997, 69.0588
+Video/A024.mp4, 10.006667, 29.97002997002997, 71.962
+Video/A025.mp4, 10.024867, 30.0, 42.0856
+Video/A026.mp4, 10.008066999999999, 30.0, 74.5397
+Video/A027.mp4, 10.006667, 29.97002997002997, 56.3388
+Video/A028.mp4, 10.023678, 29.916666666666668, 41.5304
+Video/A029.mp4, 10.004999999999999, 29.97002997002997, 75.3861
+Video/A030.mp4, 10.031478, 30.0, 81.5278
+Video/A031.mp4, 10.023522, 30.0, 62.1042
+Video/A032.mp4, 10.04, 29.97002997002997, 74.3189
+Video/A033.mp4, 10.017267, 30.0, 71.7097
+Video/A034.mp4, 10.04, 29.97002997002997, 80.4271
+Video/A035.mp4, 10.04, 29.97002997002997, 85.0968
+Video/A036.mp4, 10.028767, 30.0, 46.6608
+Video/A037.mp4, 10.04, 29.97002997002997, 77.5858
+Video/A038.mp4, 10.025063, 29.925, 51.9195
+Video/A039.mp4, 10.025599999999999, 29.916666666666668, 53.006
+Video/A040.mp4, 10.032022, 30.0, 50.3763
+Video/A041.mp4, 10.004622, 30.0, 35.1412
+Video/A042.mp4, 10.030278, 30.0, 76.8312
+Video/A043.mp4, 10.031666999999999, 30.0, 38.1554
+Video/A044.mp4, 10.033332999999999, 30.0, 53.4333
+Video/A045.mp4, 10.001510999999999, 30.0, 81.9773
+Video/A046.mp4, 10.030344, 30.0, 69.1902
+Video/A047.mp4, 10.030356, 30.0, 58.4895
+Video/A048.mp4, 10.006667, 29.97002997002997, 72.3462
+Video/A049.mp4, 10.006667, 29.97002997002997, 72.2176
+Video/A050.mp4, 10.002177999999999, 30.0, 48.0519
+Video/A051.mp4, 10.008333, 29.97002997002997, 80.2807
+Video/A052.mp4, 10.006667, 29.97002997002997, 44.8256
+Video/A053.mp4, 10.030433, 30.0, 67.0884
+Video/A054.mp4, 10.027743, 29.917, 76.1667
+Video/A055.mp4, 10.034666999999999, 30.0, 53.8092
+Video/A056.mp4, 10.0, 30.0, 52.3594
+Video/A057.mp4, 10.004999999999999, 29.97002997002997, 66.4892
+Video/A058.mp4, 10.000656, 29.850746268656717, 57.0
+Video/A059.mp4, 10.011099999999999, 30.0, 74.2873
+Video/A060.mp4, 10.008578, 30.0, 63.1839
+Video/A061.mp4, 10.002078, 30.0, 56.4521
+Video/A062.mp4, 10.026188999999999, 30.0, 47.5187
+Video/A063.mp4, 10.031378, 30.0, 44.8971
+Video/A064.mp4, 10.012844, 30.0, 49.8497
+Video/A065.mp4, 10.030622, 30.0, 81.8222
+Video/A066.mp4, 10.006667, 29.97002997002997, 74.7647
+Video/A067.mp4, 10.04, 29.97002997002997, 77.4551
+Video/A068.mp4, 10.005644, 30.0, 39.123
+Video/A069.mp4, 10.0, 30.0, 49.4444
+Video/A070.mp4, 10.006667, 29.97002997002997, 65.4608
+Video/A071.mp4, 10.006667, 29.97002997002997, 74.1071
+Video/A072.mp4, 10.08, 25.0, 20.4406
+Video/A073.mp4, 10.04, 29.97002997002997, 91.1313
+Video/A074.mp4, 10.004999999999999, 29.97002997002997, 57.3204
+Video/A075.mp4, 10.030356, 30.0, 58.1867
+Video/A076.mp4, 10.030332999999999, 30.0, 64.7835
+Video/A077.mp4, 10.04, 29.97002997002997, 74.4483
+Video/A078.mp4, 9.971478, 30.0, 38.8777
+Video/A079.mp4, 9.999977999999999, 30.0, 81.0943
+Video/A080.mp4, 10.015233, 30.0, 39.2086
+Video/A081.mp4, 10.006667, 29.97002997002997, 75.7143
+Video/A082.mp4, 10.008333, 29.97002997002997, 87.0859
+Video/A083.mp4, 10.023052999999999, 29.931, 46.3077
+Video/A084.mp4, 10.003357, 29.79, 55.3835
+Video/A085.mp4, 10.030367, 30.0, 70.8486
+Video/A086.mp4, 10.006667, 29.97002997002997, 35.2749
+Video/A087.mp4, 10.019810999999999, 29.916666666666668, 56.6415
+Video/A088.mp4, 10.027033, 29.416666666666668, 43.4892
+Video/A089.mp4, 10.019771, 29.841, 46.9444
+Video/A090.mp4, 10.006667, 29.97002997002997, 86.1709
+Video/A091.mp4, 10.021115, 29.837, 46.2238
+Video/A092.mp4, 10.026022, 29.833333333333332, 71.095
+Video/A093.mp4, 10.005956, 30.0, 66.9632
+Video/A094.mp4, 10.031467, 30.0, 52.3441
+Video/A095.mp4, 10.006667, 29.97002997002997, 71.5491
+Video/A096.mp4, 10.008333, 29.97002997002997, 66.5815
+Video/A097.mp4, 10.041667, 29.97002997002997, 66.4593
+Video/A098.mp4, 10.030356, 30.0, 85.8091
+Video/A099.mp4, 10.030510999999999, 30.0, 46.1517
+Video/A100.mp4, 10.024967, 30.0, 46.6273
+Video/A101.mp4, 10.008333, 29.97002997002997, 69.3657
+Video/A102.mp4, 10.026043999999999, 29.850746268656717, 62.987
+Video/A103.mp4, 10.016499999999999, 29.850746268656717, 71.6203
+Video/A104.mp4, 10.021355999999999, 30.0, 50.385
+Video/A105.mp4, 10.006667, 29.97002997002997, 37.8824
+Video/A106.mp4, 9.984587999999999, 29.833333333333332, 72.642
+Video/A107.mp4, 10.006667, 29.97002997002997, 81.6325
+Video/A108.mp4, 10.005556, 120.0, 35.8462
+Video/A109.mp4, 10.003021, 29.791, 36.5086
+Video/A110.mp4, 10.04, 29.97002997002997, 87.8585
+Video/B001.mp4, 10.019677999999999, 11.083333333333334, 27.2485
+Video/B002.mp4, 10.031467, 30.0, 73.4245
+Video/B003.mp4, 10.031367, 30.0, 49.1111
+Video/B004.mp4, 10.006321999999999, 30.0, 60.6719
+Video/B005.mp4, 10.038333, 29.97002997002997, 47.2959
+Video/B006.mp4, 9.998232999999999, 30.0, 44.4845
+Video/B007.mp4, 9.998267, 30.0, 37.858
+Video/B008.mp4, 10.030766999999999, 30.0, 75.0204
+Video/B009.mp4, 9.998066999999999, 30.0, 39.4971
+Video/B010.mp4, 10.002689, 29.916666666666668, 33.9518
+Video/B011.mp4, 10.031644, 29.916666666666668, 36.0333
+Video/B012.mp4, 10.024211, 29.916666666666668, 33.9801
+Video/B013.mp4, 10.025466999999999, 29.916666666666668, 58.0795
+Video/B014.mp4, 10.030267, 30.0, 62.3243
+Video/B015.mp4, 10.033332999999999, 30.0, 56.9636
+Video/B016.mp4, 10.04, 29.97002997002997, 52.1065
+Video/B017.mp4, 10.01, 29.97002997002997, 65.3693
+Video/B018.mp4, 10.001944, 30.0, 54.024
+Video/B019.mp4, 10.021422, 30.0, 66.1658
+Video/B020.mp4, 10.033332999999999, 30.0, 69.2781
+Video/B021.mp4, 10.038333, 29.97002997002997, 84.9903
+Video/B022.mp4, 10.038333, 29.97002997002997, 84.532
+Video/B023.mp4, 10.038333, 29.97002997002997, 70.8629
+Video/B024.mp4, 10.038333, 29.97002997002997, 67.6078
+Video/B025.mp4, 10.038333, 29.97002997002997, 86.53
+Video/B026.mp4, 10.04, 29.97002997002997, 72.0794
+Video/B027.mp4, 10.04, 29.97002997002997, 89.2689
+Video/B028.mp4, 10.04, 29.97002997002997, 78.8442
+Video/B029.mp4, 10.04, 29.97002997002997, 90.8788
+Video/B030.mp4, 10.04, 29.97002997002997, 23.7704
+Video/B031.mp4, 10.04, 29.97002997002997, 69.9155
+Video/B032.mp4, 10.04, 29.97002997002997, 62.1639
+Video/B033.mp4, 10.04, 29.97002997002997, 67.869
+Video/B034.mp4, 10.04, 29.97002997002997, 71.4773
+Video/B035.mp4, 10.04, 29.97002997002997, 43.9005
+Video/B036.mp4, 10.04, 29.97002997002997, 82.8776
+Video/B037.mp4, 10.04, 29.97002997002997, 47.7402
+Video/B038.mp4, 10.04, 29.97002997002997, 67.4641
+Video/B039.mp4, 10.04, 29.97002997002997, 61.2217
+Video/B040.mp4, 10.04, 29.97002997002997, 67.0729
+Video/B041.mp4, 10.04, 29.97002997002997, 45.799
+Video/B042.mp4, 10.04, 29.97002997002997, 62.8774
+Video/B043.mp4, 10.04, 29.97002997002997, 38.3
+Video/B044.mp4, 10.04, 29.97002997002997, 87.1329
+Video/B045.mp4, 10.04, 29.97002997002997, 64.1244
+Video/B046.mp4, 10.04, 29.97002997002997, 75.0229
+Video/B047.mp4, 10.04, 29.97002997002997, 63.7714
+Video/B048.mp4, 10.04, 29.97002997002997, 65.9427
+Video/B049.mp4, 10.04, 29.97002997002997, 78.5928
+Video/B050.mp4, 10.04, 29.97002997002997, 77.2019
+Video/B051.mp4, 10.04, 29.97002997002997, 46.0611
+Video/B052.mp4, 10.04, 29.97002997002997, 58.6099
+Video/B053.mp4, 10.04, 29.97002997002997, 74.9505
+Video/B054.mp4, 10.003333, 120.0, 27.2126
+Video/B055.mp4, 10.04, 29.97002997002997, 64.193
+Video/B056.mp4, 10.038333, 29.97002997002997, 62.6744
+Video/B057.mp4, 10.04, 29.97002997002997, 59.9947
+Video/B058.mp4, 10.04, 29.97002997002997, 58.367
+Video/B059.mp4, 10.04, 29.97002997002997, 52.0
+Video/B060.mp4, 10.04, 29.97002997002997, 59.4882
+Video/B061.mp4, 10.04, 29.97002997002997, 63.7011
+Video/B062.mp4, 10.04, 29.97002997002997, 56.8361
+Video/B063.mp4, 10.038333, 29.97002997002997, 44.152
+Video/B064.mp4, 10.04, 29.97002997002997, 60.0559
+Video/B065.mp4, 10.04, 29.97002997002997, 58.7207
+Video/B066.mp4, 10.04, 29.97002997002997, 57.5947
+Video/B067.mp4, 10.0, 30.0, 70.6193
+Video/B068.mp4, 10.0, 30.0, 76.6719
+Video/B069.mp4, 10.0, 30.0, 68.4948
+Video/B070.mp4, 10.0, 30.0, 56.4128
+Video/B071.mp4, 10.0, 30.0, 81.3516
+Video/B072.mp4, 10.0, 30.0, 82.573
+Video/B073.mp4, 10.0, 30.0, 66.0347
+Video/B074.mp4, 10.0, 30.0, 83.0838
+Video/B075.mp4, 10.0, 30.0, 55.4301
+Video/B076.mp4, 10.0, 30.0, 85.2071
+Video/B077.mp4, 10.0, 30.0, 81.9227
+Video/B078.mp4, 10.0, 30.0, 87.6489
+Video/B079.mp4, 10.0, 30.0, 73.5323
+Video/B080.mp4, 10.0, 30.0, 78.8136
+Video/B081.mp4, 10.033332999999999, 30.0, 46.3822
+Video/B082.mp4, 10.027854999999999, 29.916666666666668, 39.3422
+Video/B083.mp4, 10.01001, 29.97, 47.1299
+Video/B084.mp4, 10.0, 30.0, 27.3091
+Video/B085.mp4, 10.033332999999999, 30.0, 78.5556
+Video/B086.mp4, 10.033332999999999, 30.0, 80.3536
+Video/B087.mp4, 10.0, 30.0, 72.178
+Video/B088.mp4, 10.0, 30.0, 52.8177
+Video/B089.mp4, 10.01, 29.97002997002997, 74.2079
+Video/B090.mp4, 10.016807, 29.75, 26.0497
+Video/B091.mp4, 10.022345999999999, 29.833333333333332, 71.7738
+Video/B092.mp4, 10.033332999999999, 30.0, 75.2238
+Video/B093.mp4, 10.016499999999999, 29.850746268656717, 70.0112
+Video/B094.mp4, 10.033332999999999, 30.0, 72.7368
+Video/B095.mp4, 10.033332999999999, 30.0, 72.7588
+Video/B096.mp4, 10.018093, 29.846, 57.6856
+Video/B097.mp4, 10.028329, 29.416666666666668, 60.4053
+Video/B098.mp4, 10.01, 29.97002997002997, 80.199
+Video/B099.mp4, 10.022345999999999, 29.833333333333332, 69.5399
+Video/B100.mp4, 10.01, 29.97002997002997, 80.2398
+Video/B101.mp4, 10.01, 29.97002997002997, 59.2431
+Video/B102.mp4, 10.01, 29.97002997002997, 88.0219
+Video/B103.mp4, 10.033332999999999, 30.0, 25.2484
+Video/B104.mp4, 10.033332999999999, 30.0, 71.3464
+Video/B105.mp4, 10.01, 29.97002997002997, 70.9
+Video/B106.mp4, 10.01, 29.97002997002997, 57.0578
+Video/B107.mp4, 10.01, 29.97002997002997, 76.7188
+Video/B108.mp4, 10.0, 30.0, 85.7418
+Video/B109.mp4, 10.01, 29.97002997002997, 79.3295
+Video/B110.mp4, 10.01, 29.97002997002997, 77.6527
+Video/B111.mp4, 10.0, 30.0, 57.7553
+Video/B112.mp4, 10.033332999999999, 30.0, 56.3086
+Video/B113.mp4, 10.033332999999999, 30.0, 73.7705
+Video/B114.mp4, 10.01, 29.97002997002997, 85.8129
+Video/B115.mp4, 10.033332999999999, 30.0, 39.1942
+Video/B116.mp4, 10.016499999999999, 29.850746268656717, 65.6578
+Video/B117.mp4, 10.0, 30.0, 70.7436
+Video/B118.mp4, 10.041667, 24.0, 71.8769
+Video/B119.mp4, 10.033332999999999, 30.0, 69.3909
+Video/B120.mp4, 10.043367, 29.97002997002997, 82.9815
+Video/B121.mp4, 10.033332999999999, 30.0, 54.3529
+Video/B122.mp4, 10.01, 29.97002997002997, 59.2645
+Video/B123.mp4, 10.033332999999999, 30.0, 76.6535
+Video/B124.mp4, 10.033332999999999, 30.0, 55.3564
+Video/B125.mp4, 10.033332999999999, 30.0, 52.9389
+Video/B126.mp4, 10.033332999999999, 30.0, 73.5285
+Video/B127.mp4, 10.01, 29.97002997002997, 58.382
+Video/B128.mp4, 10.01, 29.97002997002997, 59.2601
+Video/B129.mp4, 10.01, 29.97002997002997, 38.2353
+Video/B130.mp4, 10.01001, 29.97, 52.7733
+Video/B131.mp4, 10.01, 29.97002997002997, 37.062
+Video/B132.mp4, 10.036451, 20.027, 42.6774
+Video/B133.mp4, 10.01, 29.97002997002997, 81.2515
+Video/B134.mp4, 10.033332999999999, 30.0, 69.35
+Video/B135.mp4, 10.043367, 29.97002997002997, 79.775
+Video/B136.mp4, 10.01, 29.97002997002997, 71.8718
+Video/B137.mp4, 10.033332999999999, 30.0, 47.1258
+Video/B138.mp4, 10.0, 30.0, 79.3155
+Video/B139.mp4, 10.033332999999999, 30.0, 68.8052
+Video/B140.mp4, 10.043367, 29.97002997002997, 68.4787
+Video/B141.mp4, 10.01, 29.97002997002997, 86.3333
+Video/B142.mp4, 10.033332999999999, 30.0, 64.9667
+Video/B143.mp4, 10.033332999999999, 30.0, 64.0795
+Video/B144.mp4, 10.01, 29.97002997002997, 46.1283
+Video/B145.mp4, 10.043367, 29.97002997002997, 87.7442
+Video/B146.mp4, 10.089385, 29.833333333333332, 62.9054
+Video/B147.mp4, 10.022345999999999, 29.833333333333332, 73.0254
+Video/B148.mp4, 10.027854999999999, 29.916666666666668, 27.5249
+Video/B149.mp4, 10.033332999999999, 30.0, 51.5759
+Video/B150.mp4, 10.016499999999999, 29.850746268656717, 58.6421
+Video/B151.mp4, 10.027854999999999, 29.916666666666668, 71.6954
+Video/B152.mp4, 10.075641, 23.919073800308954, 70.9887
+Video/B153.mp4, 10.012811, 29.662, 62.9021
+Video/B154.mp4, 10.031096, 29.90700104493208, 49.736
+Video/B155.mp4, 10.01, 29.97002997002997, 43.4969
+Video/B156.mp4, 10.0, 30.0, 69.2067
+Video/B157.mp4, 10.0, 30.0, 75.9554
+Video/B158.mp4, 10.043367, 29.97002997002997, 76.8757
+Video/B159.mp4, 10.066666999999999, 30.0, 35.9529
+Video/B160.mp4, 10.022122, 29.834, 34.3556
+Video/B161.mp4, 9.566666999999999, 30.0, 71.6517
+Video/B162.mp4, 10.050419999999999, 29.75, 51.3829
+Video/B163.mp4, 10.005035999999999, 29.785, 45.5813
+Video/B164.mp4, 10.043367, 29.97002997002997, 83.1737
+Video/B165.mp4, 10.01, 29.97002997002997, 67.3736
+Video/B166.mp4, 10.033332999999999, 30.0, 57.8166
+Video/B167.mp4, 10.0, 30.0, 14.8662
+Video/B168.mp4, 10.01, 29.97002997002997, 56.3245
+Video/B169.mp4, 10.0, 30.0, 54.0769
+Video/B170.mp4, 10.027854999999999, 29.916666666666668, 35.8132
+Video/B171.mp4, 10.01, 29.97002997002997, 89.7447
+Video/B172.mp4, 10.01, 29.97002997002997, 75.619
+Video/B173.mp4, 10.0, 30.0, 66.0171
+Video/B174.mp4, 10.0, 30.0, 17.9012
+Video/B175.mp4, 10.027854999999999, 29.916666666666668, 38.6221
+Video/B176.mp4, 10.01, 29.97002997002997, 68.9807
+Video/B177.mp4, 10.033332999999999, 30.0, 40.7351
+Video/B178.mp4, 10.033332999999999, 30.0, 61.3542
+Video/B179.mp4, 10.033332999999999, 30.0, 56.5635
+Video/B180.mp4, 10.01, 29.97002997002997, 72.9946
+Video/B181.mp4, 10.055866, 29.833333333333332, 38.5263
+Video/B182.mp4, 10.01, 29.97002997002997, 82.9684
+Video/B183.mp4, 10.043367, 29.97002997002997, 85.2941
+Video/B184.mp4, 10.033332999999999, 30.0, 57.0256
+Video/B185.mp4, 10.0, 30.0, 65.4143
+Video/B186.mp4, 10.01, 29.97002997002997, 64.5275
+Video/B187.mp4, 10.01, 29.97002997002997, 90.199
+Video/B188.mp4, 10.0, 30.0, 70.3474
+Video/B189.mp4, 10.043367, 29.97002997002997, 49.8683
+Video/B190.mp4, 10.01001, 29.97, 72.1381
+Video/B191.mp4, 10.043367, 29.97002997002997, 82.8227
+Video/B192.mp4, 10.066666999999999, 30.0, 55.6981
+Video/B193.mp4, 10.01, 29.97002997002997, 70.5926
+Video/B194.mp4, 10.0, 30.0, 55.5238
+Video/B195.mp4, 10.027854999999999, 29.916666666666668, 70.2275
+Video/B196.mp4, 10.033332999999999, 30.0, 50.7012
+Video/B197.mp4, 10.004363999999999, 29.787, 36.4242
+Video/B198.mp4, 10.043367, 29.97002997002997, 84.7438
+Video/B199.mp4, 10.01001, 29.97, 76.1538
+Video/B200.mp4, 10.01, 29.97002997002997, 36.2614
+Video/B201.mp4, 10.0, 30.0, 49.4186
+Video/B202.mp4, 10.01, 29.97002997002997, 78.8095
+Video/B203.mp4, 10.01, 29.97002997002997, 78.539
+Video/B204.mp4, 10.01, 29.97002997002997, 72.3743
+Video/B205.mp4, 10.0, 30.0, 63.2195
+Video/B206.mp4, 10.027854999999999, 29.916666666666668, 80.9784
+Video/B207.mp4, 10.01, 29.97002997002997, 86.5106
+Video/B208.mp4, 10.019802, 25.25, 74.1111
+Video/B209.mp4, 10.033332999999999, 30.0, 74.7753
+Video/B210.mp4, 10.01, 29.97002997002997, 29.4167
+Video/B211.mp4, 10.01, 29.97002997002997, 80.995
+Video/B212.mp4, 10.028329, 29.416666666666668, 71.6895
+Video/B213.mp4, 10.01, 29.97002997002997, 85.9296
+Video/B214.mp4, 10.01, 29.97002997002997, 45.2158
+Video/B215.mp4, 10.0, 30.0, 49.1789
+Video/B216.mp4, 10.033332999999999, 30.0, 85.8934
+Video/B217.mp4, 10.01, 29.97002997002997, 37.7778
+Video/B218.mp4, 10.043367, 29.97002997002997, 89.0446
+Video/B219.mp4, 10.043367, 29.97002997002997, 78.0149
+Video/B220.mp4, 10.01, 29.97002997002997, 70.117
+Video/B221.mp4, 10.0, 30.0, 80.4144
+Video/B222.mp4, 10.083333, 24.0, 34.2513
+Video/B223.mp4, 10.01, 29.97002997002997, 60.6067
+Video/B224.mp4, 10.01, 29.97002997002997, 81.0055
+Video/B225.mp4, 10.033332999999999, 30.0, 80.4624
+Video/B226.mp4, 10.022345999999999, 29.833333333333332, 62.5157
+Video/B227.mp4, 10.01, 29.97002997002997, 90.2448
+Video/B228.mp4, 10.033332999999999, 30.0, 71.1987
+Video/B229.mp4, 10.01, 29.97002997002997, 73.8866
+Video/B230.mp4, 10.03009, 29.91, 73.6337
+Video/B231.mp4, 10.01, 29.97002997002997, 65.0412
+Video/B232.mp4, 10.0, 29.8, 65.1549
+Video/B233.mp4, 10.016499999999999, 29.850746268656717, 50.8201
+Video/B234.mp4, 10.033332999999999, 30.0, 80.4612
+Video/B235.mp4, 10.0, 30.0, 68.4103
+Video/B236.mp4, 10.020107, 29.84, 53.3737
+Video/B237.mp4, 10.033332999999999, 30.0, 70.1209
+Video/B238.mp4, 10.016499999999999, 29.850746268656717, 60.5058
+Video/B239.mp4, 10.0, 30.0, 67.5337
+Video/B240.mp4, 10.086454999999999, 29.842, 67.1576
+Video/B241.mp4, 10.027854999999999, 29.916666666666668, 66.459
+Video/B242.mp4, 10.033332999999999, 30.0, 71.6724
+Video/B243.mp4, 10.016499999999999, 29.850746268656717, 47.8021
+Video/B244.mp4, 10.033332999999999, 30.0, 63.5397
+Video/B245.mp4, 10.033332999999999, 30.0, 82.8444
+Video/B246.mp4, 10.033332999999999, 30.0, 56.3813
+Video/B247.mp4, 10.033332999999999, 30.0, 76.2626
+Video/B248.mp4, 10.01, 29.97002997002997, 62.3922
+Video/B249.mp4, 10.0, 30.0, 69.7869
+Video/B250.mp4, 10.016499999999999, 29.850746268656717, 61.2128
+Video/B251.mp4, 10.01, 29.97002997002997, 82.7214
+Video/B252.mp4, 10.015691, 29.953, 73.2762
+Video/B253.mp4, 10.033332999999999, 30.0, 72.1818
+Video/B254.mp4, 10.033332999999999, 30.0, 54.569
+Video/B255.mp4, 10.022345999999999, 29.833333333333332, 62.2424
+Video/B256.mp4, 10.0, 30.0, 59.5503
+Video/B257.mp4, 10.01, 29.97002997002997, 82.2251
+Video/B258.mp4, 10.01, 29.97002997002997, 85.4278
+Video/B259.mp4, 10.01, 29.97002997002997, 84.8989
+Video/B260.mp4, 10.041667, 24.0, 6.22368
+Video/B261.mp4, 10.033332999999999, 30.0, 78.0125
+Video/B262.mp4, 10.01, 29.97002997002997, 51.9461
+Video/B263.mp4, 10.033332999999999, 30.0, 64.4931
+Video/B264.mp4, 10.033332999999999, 30.0, 78.114
+Video/B265.mp4, 9.7, 30.0, 74.4638
+Video/B266.mp4, 10.01, 29.97002997002997, 88.24
+Video/B267.mp4, 10.01, 29.97002997002997, 62.3061
+Video/B268.mp4, 10.027854999999999, 29.916666666666668, 25.0861
+Video/B269.mp4, 10.022345999999999, 29.833333333333332, 66.5404
+Video/B270.mp4, 10.043367, 29.97002997002997, 90.3281
+Video/B271.mp4, 10.0, 30.0, 84.2073
+Video/B272.mp4, 10.041667, 24.0, 49.6404
+Video/B273.mp4, 10.066666999999999, 30.0, 53.4568
+Video/B274.mp4, 10.033332999999999, 30.0, 54.9497
+Video/B275.mp4, 10.043367, 29.97002997002997, 79.0355
+Video/B276.mp4, 10.043367, 29.97002997002997, 83.5027
+Video/B277.mp4, 10.01, 29.97002997002997, 28.6053
+Video/B278.mp4, 10.01, 29.97002997002997, 70.3646
+Video/B279.mp4, 10.033332999999999, 30.0, 67.9827
+Video/B280.mp4, 10.0, 30.0, 69.5136
+Video/B281.mp4, 10.01, 29.97002997002997, 50.288
+Video/B282.mp4, 9.994429, 29.916666666666668, 70.8817
+Video/B283.mp4, 10.027009, 29.62, 64.2865
+Video/B284.mp4, 10.01, 29.97002997002997, 71.3248
+Video/B285.mp4, 10.01, 29.97002997002997, 73.6837
+Video/B286.mp4, 10.033332999999999, 30.0, 38.6012
+Video/B287.mp4, 10.0, 30.0, 70.2771
+Video/B288.mp4, 10.043367, 29.97002997002997, 82.4922
+Video/B289.mp4, 10.01, 29.97002997002997, 84.8406
+Video/B290.mp4, 10.033332999999999, 30.0, 62.4348
+Video/B291.mp4, 10.0, 30.0, 71.1256
+Video/B292.mp4, 10.033332999999999, 30.0, 81.9686
+Video/B293.mp4, 10.066666999999999, 30.0, 78.2485
+Video/B294.mp4, 10.0, 30.0, 66.3431
+Video/B295.mp4, 10.01, 29.97002997002997, 54.7539
+Video/B296.mp4, 10.021044, 29.937, 63.203
+Video/B297.mp4, 10.033332999999999, 30.0, 61.4341
+Video/B298.mp4, 10.033332999999999, 30.0, 42.3916
+Video/B299.mp4, 10.033332999999999, 30.0, 72.21
+Video/B300.mp4, 10.033332999999999, 30.0, 27.691
+Video/B301.mp4, 10.0, 30.0, 49.1242
+Video/B302.mp4, 10.01, 29.97002997002997, 72.25
+Video/B303.mp4, 10.041667, 24.0, 18.8824
+Video/B304.mp4, 10.043367, 29.97002997002997, 91.7312
+Video/B305.mp4, 10.01, 29.97002997002997, 66.0914
+Video/B306.mp4, 10.033332999999999, 30.0, 79.1263
+Video/B307.mp4, 10.022345999999999, 29.833333333333332, 62.3543
+Video/B308.mp4, 10.033332999999999, 30.0, 82.8098
+Video/B309.mp4, 10.0, 30.0, 75.774
+Video/B310.mp4, 10.033332999999999, 30.0, 85.1746
+Video/B311.mp4, 10.033332999999999, 30.0, 88.8258
+Video/B312.mp4, 10.033332999999999, 30.0, 61.1696
+Video/B313.mp4, 10.033332999999999, 30.0, 89.2074
+Video/B314.mp4, 10.01, 29.97002997002997, 94.2865
+Video/B315.mp4, 10.033332999999999, 30.0, 85.593
+Video/B316.mp4, 10.033332999999999, 30.0, 89.1236
+Video/C001.mp4, 10.04, 30.0, 84.2191
+Video/D001.mp4, 10.020211, 16.666666666666668, 32.7293
+Video/E001.mp4, 10.0314, 30.0, 51.6061
+Video/F001.mp4, 10.026644, 30.020013342228154, 69.3289
+Video/F002.mp4, 10.007033, 30.0, 29.0101
+Video/F003.mp4, 10.011688999999999, 30.0, 40.2153
+Video/F004.mp4, 10.001610999999999, 30.0, 18.6163
+Video/F005.mp4, 10.004222, 30.0, 56.6974
+Video/F006.mp4, 10.01, 29.97002997002997, 59.6044
+Video/F007.mp4, 10.008333, 29.97002997002997, 42.7267
+Video/G001.mp4, 10.0, 30.0, 72.0884
+Video/G002.mp4, 10.033332999999999, 30.0, 47.5172
+Video/G003.mp4, 10.033332999999999, 30.0, 73.1593
+Video/G004.mp4, 10.0, 30.0, 76.068
+Video/G005.mp4, 10.033332999999999, 30.0, 85.2933
+Video/G006.mp4, 10.0, 30.0, 72.881
+Video/G007.mp4, 10.01, 29.97002997002997, 87.2169
+Video/G008.mp4, 10.01, 29.97002997002997, 65.8113
+Video/G009.mp4, 10.01, 29.97002997002997, 72.7931
+Video/G010.mp4, 10.01, 29.97002997002997, 68.4129
+Video/G011.mp4, 10.01, 29.97002997002997, 24.7697
+Video/G012.mp4, 10.01, 29.97002997002997, 51.3923
+Video/G013.mp4, 10.036451, 20.027, 73.2679
+Video/G014.mp4, 10.01, 29.97002997002997, 56.096
+Video/G015.mp4, 10.01, 29.97002997002997, 89.2513
+Video/G016.mp4, 10.01, 29.97002997002997, 80.5959
+Video/G017.mp4, 10.0, 30.0, 75.1832
+Video/G018.mp4, 10.0, 30.0, 83.5848
+Video/G019.mp4, 10.0, 30.0, 71.6599
+Video/G020.mp4, 10.033332999999999, 30.0, 60.6335
+Video/G021.mp4, 10.0, 30.0, 41.2703
+Video/G022.mp4, 10.0, 30.0, 58.5622
+Video/G023.mp4, 10.0, 30.0, 58.9406
+Video/G024.mp4, 10.033332999999999, 30.0, 59.6292
+Video/G025.mp4, 10.0, 30.0, 40.5028
+Video/G026.mp4, 10.033332999999999, 30.0, 54.4118
+Video/G027.mp4, 10.033332999999999, 30.0, 45.7368
+Video/G028.mp4, 10.033332999999999, 30.0, 61.3883
+Video/G029.mp4, 10.033332999999999, 30.0, 50.4835
+Video/G030.mp4, 10.01, 29.97002997002997, 77.7895
+Video/G031.mp4, 10.01, 29.97002997002997, 70.6823
+Video/G032.mp4, 10.01, 29.97002997002997, 53.278
+Video/G033.mp4, 10.01, 29.97002997002997, 60.6505
+Video/G034.mp4, 10.01, 29.97002997002997, 82.4354
+Video/G035.mp4, 10.01, 29.97002997002997, 72.2569
+Video/G036.mp4, 10.01, 29.97002997002997, 78.6898
+Video/G037.mp4, 10.01, 29.97002997002997, 78.2783
+Video/G038.mp4, 10.01, 29.97002997002997, 72.125
+Video/G039.mp4, 10.01, 29.97002997002997, 73.5789
+Video/G040.mp4, 10.01, 29.97002997002997, 80.9306
+Video/G041.mp4, 10.016499999999999, 29.850746268656717, 43.7181
+Video/G042.mp4, 10.022345999999999, 29.833333333333332, 70.191
+Video/G043.mp4, 10.022345999999999, 29.833333333333332, 77.4279
+Video/G044.mp4, 10.036432999999999, 29.791459781529294, 74.9635
+Video/G045.mp4, 10.016499999999999, 29.850746268656717, 62.4847
+Video/G046.mp4, 10.016499999999999, 29.850746268656717, 78.8212
+Video/G047.mp4, 10.066666999999999, 30.0, 70.1804
+Video/G048.mp4, 10.066666999999999, 30.0, 78.9434
+Video/G049.mp4, 9.994429, 29.916666666666668, 44.5829
+Video/G050.mp4, 10.033332999999999, 30.0, 79.9171
+Video/G051.mp4, 10.066666999999999, 30.0, 75.6131
+Video/G052.mp4, 10.066666999999999, 30.0, 67.6825
+Video/G053.mp4, 10.033332999999999, 30.0, 74.3118
+Video/G054.mp4, 10.0, 30.0, 56.0
+Video/G055.mp4, 10.066666999999999, 30.0, 63.478
+Video/G056.mp4, 10.01, 29.97002997002997, 62.0506
+Video/G057.mp4, 10.033332999999999, 30.0, 54.5354
+Video/G058.mp4, 10.033332999999999, 30.0, 65.1421
+Video/G059.mp4, 10.033332999999999, 30.0, 38.0054
+Video/G060.mp4, 10.033332999999999, 30.0, 71.4348
+Video/G061.mp4, 10.033332999999999, 30.0, 73.8883
+Video/G062.mp4, 10.041667, 24.0, 49.3832
+Video/G063.mp4, 10.115701999999999, 20.166666666666668, 29.0845
+Video/G064.mp4, 10.016529, 20.166666666666668, 11.3333
+Video/G065.mp4, 10.020619, 24.25, 68.6814
+Video/G066.mp4, 10.097999999999999, 20.2020202020202, 11.9079
+Video/G067.mp4, 10.022345999999999, 29.833333333333332, 57.2944
+Video/G068.mp4, 10.016499999999999, 29.850746268656717, 59.4638
+Video/G069.mp4, 10.022345999999999, 29.833333333333332, 66.1244
+Video/G070.mp4, 10.066666999999999, 30.0, 79.5904
+Video/G071.mp4, 10.0, 30.0, 48.3714
+Video/G072.mp4, 10.01, 29.97002997002997, 47.7432
+Video/G073.mp4, 10.0, 30.0, 18.6333
+Video/G074.mp4, 10.01, 29.97002997002997, 83.1826
+Video/G075.mp4, 10.049915, 29.851, 64.293
+Video/G076.mp4, 10.020443, 29.839, 80.0361
+Video/G077.mp4, 10.022345999999999, 29.833333333333332, 85.2383
+Video/G078.mp4, 10.022345999999999, 29.833333333333332, 81.5829
+Video/G079.mp4, 10.033332999999999, 30.0, 15.3576
+Video/G080.mp4, 10.027854999999999, 29.916666666666668, 26.2575
+Video/G081.mp4, 10.041667, 24.0, 46.7325
+Video/G082.mp4, 10.043367, 29.97002997002997, 78.019
+Video/G083.mp4, 10.01, 29.97002997002997, 78.7753
+Video/G084.mp4, 10.0, 30.0, 59.3198
+Video/G085.mp4, 10.01, 29.97002997002997, 76.4971
+Video/G086.mp4, 10.01, 29.97002997002997, 41.3556
+Video/G087.mp4, 10.01, 29.97002997002997, 71.1184
+Video/G088.mp4, 10.01, 29.97002997002997, 56.0826
+Video/G089.mp4, 10.041667, 24.0, 31.2638
+Video/G090.mp4, 10.043367, 29.97002997002997, 75.4464
+Video/G091.mp4, 10.01, 29.97002997002997, 84.2475
+Video/G092.mp4, 10.01, 29.97002997002997, 67.8351
+Video/G093.mp4, 10.01, 29.97002997002997, 66.1105
+Video/G094.mp4, 10.01, 29.97002997002997, 66.865
+Video/G095.mp4, 10.01, 29.97002997002997, 79.2192
+Video/G096.mp4, 10.01, 29.97002997002997, 86.6243
+Video/G097.mp4, 10.01, 29.97002997002997, 84.6976
+Video/G098.mp4, 10.01, 29.97002997002997, 35.8556
+Video/G099.mp4, 10.01, 29.97002997002997, 80.5632
+Video/G100.mp4, 10.0, 30.0, 80.0573
+Video/G101.mp4, 10.01, 29.97002997002997, 63.0345
+Video/G102.mp4, 10.01, 29.97002997002997, 67.75
+Video/G103.mp4, 10.01, 29.97002997002997, 81.3545
+Video/G104.mp4, 10.01, 29.97002997002997, 86.1474
+Video/G105.mp4, 10.033332999999999, 30.0, 45.0703
+Video/G106.mp4, 10.043367, 29.97002997002997, 23.4251
+Video/G107.mp4, 10.033332999999999, 30.0, 49.2717
+Video/G108.mp4, 10.041667, 24.0, 58.5988
+Video/G109.mp4, 10.0, 30.0, 79.2459
+Video/G110.mp4, 10.0, 30.0, 83.5217
+Video/G111.mp4, 10.033332999999999, 30.0, 55.0282
+Video/G112.mp4, 10.1, 30.0, 74.3393
+Video/G113.mp4, 10.043367, 29.97002997002997, 74.4923
+Video/G114.mp4, 10.033332999999999, 30.0, 65.0532
+Video/G115.mp4, 10.043367, 29.97002997002997, 77.8551
+Video/G116.mp4, 10.043367, 29.97002997002997, 88.4254
+Video/G117.mp4, 10.043367, 29.97002997002997, 71.7751
+Video/G118.mp4, 10.043367, 29.97002997002997, 65.9
+Video/G119.mp4, 10.043367, 29.97002997002997, 76.8795
+Video/H001.mp4, 10.0, 30.0, 63.6974
+Video/H002.mp4, 10.010714, 29.868, 69.1073
+Video/I001.mp4, 10.01, 29.97002997002997, 72.3462
+Video/I002.mp4, 10.022122, 29.834, 63.4225
+Video/I003.mp4, 10.018032, 29.946, 69.5414
+Video/J001.mp4, 10.011974, 30.064, 71.2778
+Video/J002.mp4, 10.022197, 29.734, 62.36
+Video/J003.mp4, 10.003988, 30.088, 60.4424
+Video/J004.mp4, 10.028224999999999, 30.115, 62.645
+Video/J005.mp4, 10.034182, 27.207, 24.1117
+Video/K001.mp4, 10.0, 30.0, 33.0114
+Video/K002.mp4, 10.016694, 29.95, 53.9415
+Video/L001.mp4, 10.006667, 29.97002997002997, 76.5463
+Video/M001.mp4, 10.0, 30.0, 36.9665
+Video/M002.mp4, 10.0, 30.0, 69.3081
+Video/M003.mp4, 10.0, 30.0, 21.1084
+Video/N001.mp4, 10.004999999999999, 120.0, 67.445
+Video/O001.mp4, 10.036667, 24.0, 27.8
+Video/O002.mp4, 10.04, 120.0, 31.0169
+Video/P001.mp4, 10.0, 30.0, 64.399
+Video/P002.mp4, 10.0, 30.0, 60.1489
+Video/P003.mp4, 10.0, 30.0, 61.749
+Video/P004.mp4, 10.0, 30.0, 63.4194
+Video/P005.mp4, 10.0, 30.0, 74.1421
+Video/P006.mp4, 10.0, 30.0, 54.2271
+Video/P007.mp4, 10.0, 30.0, 38.2139
+Video/P008.mp4, 10.0, 30.0, 38.4545
+Video/P009.mp4, 10.0, 30.0, 55.9663
+Video/Q001.mp4, 10.026644, 30.020013342228154, 61.1466
+Video/R001.mp4, 10.026644, 30.020013342228154, 72.4848

examplar_data_labels/PIPAL_NTIRE22/labels.txt

@@ -0,0 +1,1650 @@
+A0000_10_00.bmp, -1, -1, -1
+A0000_10_01.bmp, -1, -1, -1
+A0000_10_02.bmp, -1, -1, -1
+A0000_10_03.bmp, -1, -1, -1
+A0000_10_04.bmp, -1, -1, -1
+A0000_10_05.bmp, -1, -1, -1
+A0000_10_06.bmp, -1, -1, -1
+A0000_10_07.bmp, -1, -1, -1
+A0000_10_08.bmp, -1, -1, -1
+A0000_10_10.bmp, -1, -1, -1
+A0000_10_11.bmp, -1, -1, -1
+A0000_10_12.bmp, -1, -1, -1
+A0000_10_14.bmp, -1, -1, -1
+A0000_10_15.bmp, -1, -1, -1
+A0000_10_16.bmp, -1, -1, -1
+A0000_10_17.bmp, -1, -1, -1
+A0000_10_18.bmp, -1, -1, -1
+A0000_10_19.bmp, -1, -1, -1
+A0000_10_20.bmp, -1, -1, -1
+A0000_10_21.bmp, -1, -1, -1
+A0000_10_24.bmp, -1, -1, -1
+A0000_10_26.bmp, -1, -1, -1
+A0000_10_27.bmp, -1, -1, -1
+A0000_10_28.bmp, -1, -1, -1
+A0000_10_29.bmp, -1, -1, -1
+A0000_10_30.bmp, -1, -1, -1
+A0000_10_31.bmp, -1, -1, -1
+A0000_10_32.bmp, -1, -1, -1
+A0000_10_34.bmp, -1, -1, -1
+A0000_10_35.bmp, -1, -1, -1
+A0000_10_36.bmp, -1, -1, -1
+A0000_10_37.bmp, -1, -1, -1
+A0000_10_38.bmp, -1, -1, -1
+A0000_10_39.bmp, -1, -1, -1
+A0000_10_40.bmp, -1, -1, -1
+A0000_10_41.bmp, -1, -1, -1
+A0000_10_42.bmp, -1, -1, -1
+A0000_10_43.bmp, -1, -1, -1
+A0000_10_44.bmp, -1, -1, -1
+A0000_10_46.bmp, -1, -1, -1
+A0000_10_47.bmp, -1, -1, -1
+A0000_10_48.bmp, -1, -1, -1
+A0000_10_49.bmp, -1, -1, -1
+A0000_10_50.bmp, -1, -1, -1
+A0000_10_51.bmp, -1, -1, -1
+A0000_10_52.bmp, -1, -1, -1
+A0000_10_53.bmp, -1, -1, -1
+A0000_10_54.bmp, -1, -1, -1
+A0000_10_55.bmp, -1, -1, -1
+A0000_10_56.bmp, -1, -1, -1
+A0000_10_58.bmp, -1, -1, -1
+A0000_10_61.bmp, -1, -1, -1
+A0000_10_62.bmp, -1, -1, -1
+A0000_10_63.bmp, -1, -1, -1
+A0000_10_64.bmp, -1, -1, -1
+A0000_10_65.bmp, -1, -1, -1
+A0000_10_66.bmp, -1, -1, -1
+A0000_10_67.bmp, -1, -1, -1
+A0000_10_68.bmp, -1, -1, -1
+A0000_10_69.bmp, -1, -1, -1
+A0000_10_70.bmp, -1, -1, -1
+A0000_10_71.bmp, -1, -1, -1
+A0000_10_72.bmp, -1, -1, -1
+A0000_10_73.bmp, -1, -1, -1
+A0000_10_74.bmp, -1, -1, -1
+A0000_10_75.bmp, -1, -1, -1
+A0004_10_00.bmp, -1, -1, -1
+A0004_10_01.bmp, -1, -1, -1
+A0004_10_02.bmp, -1, -1, -1
+A0004_10_03.bmp, -1, -1, -1
+A0004_10_04.bmp, -1, -1, -1
+A0004_10_05.bmp, -1, -1, -1
+A0004_10_06.bmp, -1, -1, -1
+A0004_10_07.bmp, -1, -1, -1
+A0004_10_08.bmp, -1, -1, -1
+A0004_10_10.bmp, -1, -1, -1
+A0004_10_11.bmp, -1, -1, -1
+A0004_10_12.bmp, -1, -1, -1
+A0004_10_14.bmp, -1, -1, -1
+A0004_10_15.bmp, -1, -1, -1
+A0004_10_16.bmp, -1, -1, -1
+A0004_10_17.bmp, -1, -1, -1
+A0004_10_18.bmp, -1, -1, -1
+A0004_10_19.bmp, -1, -1, -1
+A0004_10_20.bmp, -1, -1, -1
+A0004_10_21.bmp, -1, -1, -1
+A0004_10_24.bmp, -1, -1, -1
+A0004_10_26.bmp, -1, -1, -1
+A0004_10_27.bmp, -1, -1, -1
+A0004_10_28.bmp, -1, -1, -1
+A0004_10_29.bmp, -1, -1, -1
+A0004_10_30.bmp, -1, -1, -1
+A0004_10_31.bmp, -1, -1, -1
+A0004_10_32.bmp, -1, -1, -1
+A0004_10_34.bmp, -1, -1, -1
+A0004_10_35.bmp, -1, -1, -1
+A0004_10_36.bmp, -1, -1, -1
+A0004_10_37.bmp, -1, -1, -1
+A0004_10_38.bmp, -1, -1, -1
+A0004_10_39.bmp, -1, -1, -1
+A0004_10_40.bmp, -1, -1, -1
+A0004_10_41.bmp, -1, -1, -1
+A0004_10_42.bmp, -1, -1, -1
+A0004_10_43.bmp, -1, -1, -1
+A0004_10_44.bmp, -1, -1, -1
+A0004_10_46.bmp, -1, -1, -1
+A0004_10_47.bmp, -1, -1, -1
+A0004_10_48.bmp, -1, -1, -1
+A0004_10_49.bmp, -1, -1, -1
+A0004_10_50.bmp, -1, -1, -1
+A0004_10_51.bmp, -1, -1, -1
+A0004_10_52.bmp, -1, -1, -1
+A0004_10_53.bmp, -1, -1, -1
+A0004_10_54.bmp, -1, -1, -1
+A0004_10_55.bmp, -1, -1, -1
+A0004_10_56.bmp, -1, -1, -1
+A0004_10_58.bmp, -1, -1, -1
+A0004_10_61.bmp, -1, -1, -1
+A0004_10_62.bmp, -1, -1, -1
+A0004_10_63.bmp, -1, -1, -1
+A0004_10_64.bmp, -1, -1, -1
+A0004_10_65.bmp, -1, -1, -1
+A0004_10_66.bmp, -1, -1, -1
+A0004_10_67.bmp, -1, -1, -1
+A0004_10_68.bmp, -1, -1, -1
+A0004_10_69.bmp, -1, -1, -1
+A0004_10_70.bmp, -1, -1, -1
+A0004_10_71.bmp, -1, -1, -1
+A0004_10_72.bmp, -1, -1, -1
+A0004_10_73.bmp, -1, -1, -1
+A0004_10_74.bmp, -1, -1, -1
+A0004_10_75.bmp, -1, -1, -1
+A0011_10_00.bmp, -1, -1, -1
+A0011_10_01.bmp, -1, -1, -1
+A0011_10_02.bmp, -1, -1, -1
+A0011_10_03.bmp, -1, -1, -1
+A0011_10_04.bmp, -1, -1, -1
+A0011_10_05.bmp, -1, -1, -1
+A0011_10_06.bmp, -1, -1, -1
+A0011_10_07.bmp, -1, -1, -1
+A0011_10_08.bmp, -1, -1, -1
+A0011_10_10.bmp, -1, -1, -1
+A0011_10_11.bmp, -1, -1, -1
+A0011_10_12.bmp, -1, -1, -1
+A0011_10_14.bmp, -1, -1, -1
+A0011_10_15.bmp, -1, -1, -1
+A0011_10_16.bmp, -1, -1, -1
+A0011_10_17.bmp, -1, -1, -1
+A0011_10_18.bmp, -1, -1, -1
+A0011_10_19.bmp, -1, -1, -1
+A0011_10_20.bmp, -1, -1, -1
+A0011_10_21.bmp, -1, -1, -1
+A0011_10_24.bmp, -1, -1, -1
+A0011_10_26.bmp, -1, -1, -1
+A0011_10_27.bmp, -1, -1, -1
+A0011_10_28.bmp, -1, -1, -1
+A0011_10_29.bmp, -1, -1, -1
+A0011_10_30.bmp, -1, -1, -1
+A0011_10_31.bmp, -1, -1, -1
+A0011_10_32.bmp, -1, -1, -1
+A0011_10_34.bmp, -1, -1, -1
+A0011_10_35.bmp, -1, -1, -1
+A0011_10_36.bmp, -1, -1, -1
+A0011_10_37.bmp, -1, -1, -1
+A0011_10_38.bmp, -1, -1, -1
+A0011_10_39.bmp, -1, -1, -1
+A0011_10_40.bmp, -1, -1, -1
+A0011_10_41.bmp, -1, -1, -1
+A0011_10_42.bmp, -1, -1, -1
+A0011_10_43.bmp, -1, -1, -1
+A0011_10_44.bmp, -1, -1, -1
+A0011_10_46.bmp, -1, -1, -1
+A0011_10_47.bmp, -1, -1, -1
+A0011_10_48.bmp, -1, -1, -1
+A0011_10_49.bmp, -1, -1, -1
+A0011_10_50.bmp, -1, -1, -1
+A0011_10_51.bmp, -1, -1, -1
+A0011_10_52.bmp, -1, -1, -1
+A0011_10_53.bmp, -1, -1, -1
+A0011_10_54.bmp, -1, -1, -1
+A0011_10_55.bmp, -1, -1, -1
+A0011_10_56.bmp, -1, -1, -1
+A0011_10_58.bmp, -1, -1, -1
+A0011_10_61.bmp, -1, -1, -1
+A0011_10_62.bmp, -1, -1, -1
+A0011_10_63.bmp, -1, -1, -1
+A0011_10_64.bmp, -1, -1, -1
+A0011_10_65.bmp, -1, -1, -1
+A0011_10_66.bmp, -1, -1, -1
+A0011_10_67.bmp, -1, -1, -1
+A0011_10_68.bmp, -1, -1, -1
+A0011_10_69.bmp, -1, -1, -1
+A0011_10_70.bmp, -1, -1, -1
+A0011_10_71.bmp, -1, -1, -1
+A0011_10_72.bmp, -1, -1, -1
+A0011_10_73.bmp, -1, -1, -1
+A0011_10_74.bmp, -1, -1, -1
+A0011_10_75.bmp, -1, -1, -1
+A0022_10_00.bmp, -1, -1, -1
+A0022_10_01.bmp, -1, -1, -1
+A0022_10_02.bmp, -1, -1, -1
+A0022_10_03.bmp, -1, -1, -1
+A0022_10_04.bmp, -1, -1, -1
+A0022_10_05.bmp, -1, -1, -1
+A0022_10_06.bmp, -1, -1, -1
+A0022_10_07.bmp, -1, -1, -1
+A0022_10_08.bmp, -1, -1, -1
+A0022_10_10.bmp, -1, -1, -1
+A0022_10_11.bmp, -1, -1, -1
+A0022_10_12.bmp, -1, -1, -1
+A0022_10_14.bmp, -1, -1, -1
+A0022_10_15.bmp, -1, -1, -1
+A0022_10_16.bmp, -1, -1, -1
+A0022_10_17.bmp, -1, -1, -1
+A0022_10_18.bmp, -1, -1, -1
+A0022_10_19.bmp, -1, -1, -1
+A0022_10_20.bmp, -1, -1, -1
+A0022_10_21.bmp, -1, -1, -1
+A0022_10_24.bmp, -1, -1, -1
+A0022_10_26.bmp, -1, -1, -1
+A0022_10_27.bmp, -1, -1, -1
+A0022_10_28.bmp, -1, -1, -1
+A0022_10_29.bmp, -1, -1, -1
+A0022_10_30.bmp, -1, -1, -1
+A0022_10_31.bmp, -1, -1, -1
+A0022_10_32.bmp, -1, -1, -1
+A0022_10_34.bmp, -1, -1, -1
+A0022_10_35.bmp, -1, -1, -1
+A0022_10_36.bmp, -1, -1, -1
+A0022_10_37.bmp, -1, -1, -1
+A0022_10_38.bmp, -1, -1, -1
+A0022_10_39.bmp, -1, -1, -1
+A0022_10_40.bmp, -1, -1, -1
+A0022_10_41.bmp, -1, -1, -1
+A0022_10_42.bmp, -1, -1, -1
+A0022_10_43.bmp, -1, -1, -1
+A0022_10_44.bmp, -1, -1, -1
+A0022_10_46.bmp, -1, -1, -1
+A0022_10_47.bmp, -1, -1, -1
+A0022_10_48.bmp, -1, -1, -1
+A0022_10_49.bmp, -1, -1, -1
+A0022_10_50.bmp, -1, -1, -1
+A0022_10_51.bmp, -1, -1, -1
+A0022_10_52.bmp, -1, -1, -1
+A0022_10_53.bmp, -1, -1, -1
+A0022_10_54.bmp, -1, -1, -1
+A0022_10_55.bmp, -1, -1, -1
+A0022_10_56.bmp, -1, -1, -1
+A0022_10_58.bmp, -1, -1, -1
+A0022_10_61.bmp, -1, -1, -1
+A0022_10_62.bmp, -1, -1, -1
+A0022_10_63.bmp, -1, -1, -1
+A0022_10_64.bmp, -1, -1, -1
+A0022_10_65.bmp, -1, -1, -1
+A0022_10_66.bmp, -1, -1, -1
+A0022_10_67.bmp, -1, -1, -1
+A0022_10_68.bmp, -1, -1, -1
+A0022_10_69.bmp, -1, -1, -1
+A0022_10_70.bmp, -1, -1, -1
+A0022_10_71.bmp, -1, -1, -1
+A0022_10_72.bmp, -1, -1, -1
+A0022_10_73.bmp, -1, -1, -1
+A0022_10_74.bmp, -1, -1, -1
+A0022_10_75.bmp, -1, -1, -1
+A0033_10_00.bmp, -1, -1, -1
+A0033_10_01.bmp, -1, -1, -1
+A0033_10_02.bmp, -1, -1, -1
+A0033_10_03.bmp, -1, -1, -1
+A0033_10_04.bmp, -1, -1, -1
+A0033_10_05.bmp, -1, -1, -1
+A0033_10_06.bmp, -1, -1, -1
+A0033_10_07.bmp, -1, -1, -1
+A0033_10_08.bmp, -1, -1, -1
+A0033_10_10.bmp, -1, -1, -1
+A0033_10_11.bmp, -1, -1, -1
+A0033_10_12.bmp, -1, -1, -1
+A0033_10_14.bmp, -1, -1, -1
+A0033_10_15.bmp, -1, -1, -1
+A0033_10_16.bmp, -1, -1, -1
+A0033_10_17.bmp, -1, -1, -1
+A0033_10_18.bmp, -1, -1, -1
+A0033_10_19.bmp, -1, -1, -1
+A0033_10_20.bmp, -1, -1, -1
+A0033_10_21.bmp, -1, -1, -1
+A0033_10_24.bmp, -1, -1, -1
+A0033_10_26.bmp, -1, -1, -1
+A0033_10_27.bmp, -1, -1, -1
+A0033_10_28.bmp, -1, -1, -1
+A0033_10_29.bmp, -1, -1, -1
+A0033_10_30.bmp, -1, -1, -1
+A0033_10_31.bmp, -1, -1, -1
+A0033_10_32.bmp, -1, -1, -1
+A0033_10_34.bmp, -1, -1, -1
+A0033_10_35.bmp, -1, -1, -1
+A0033_10_36.bmp, -1, -1, -1
+A0033_10_37.bmp, -1, -1, -1
+A0033_10_38.bmp, -1, -1, -1
+A0033_10_39.bmp, -1, -1, -1
+A0033_10_40.bmp, -1, -1, -1
+A0033_10_41.bmp, -1, -1, -1
+A0033_10_42.bmp, -1, -1, -1
+A0033_10_43.bmp, -1, -1, -1
+A0033_10_44.bmp, -1, -1, -1
+A0033_10_46.bmp, -1, -1, -1
+A0033_10_47.bmp, -1, -1, -1
+A0033_10_48.bmp, -1, -1, -1
+A0033_10_49.bmp, -1, -1, -1
+A0033_10_50.bmp, -1, -1, -1
+A0033_10_51.bmp, -1, -1, -1
+A0033_10_52.bmp, -1, -1, -1
+A0033_10_53.bmp, -1, -1, -1
+A0033_10_54.bmp, -1, -1, -1
+A0033_10_55.bmp, -1, -1, -1
+A0033_10_56.bmp, -1, -1, -1
+A0033_10_58.bmp, -1, -1, -1
+A0033_10_61.bmp, -1, -1, -1
+A0033_10_62.bmp, -1, -1, -1
+A0033_10_63.bmp, -1, -1, -1
+A0033_10_64.bmp, -1, -1, -1
+A0033_10_65.bmp, -1, -1, -1
+A0033_10_66.bmp, -1, -1, -1
+A0033_10_67.bmp, -1, -1, -1
+A0033_10_68.bmp, -1, -1, -1
+A0033_10_69.bmp, -1, -1, -1
+A0033_10_70.bmp, -1, -1, -1
+A0033_10_71.bmp, -1, -1, -1
+A0033_10_72.bmp, -1, -1, -1
+A0033_10_73.bmp, -1, -1, -1
+A0033_10_74.bmp, -1, -1, -1
+A0033_10_75.bmp, -1, -1, -1
+A0036_10_00.bmp, -1, -1, -1
+A0036_10_01.bmp, -1, -1, -1
+A0036_10_02.bmp, -1, -1, -1
+A0036_10_03.bmp, -1, -1, -1
+A0036_10_04.bmp, -1, -1, -1
+A0036_10_05.bmp, -1, -1, -1
+A0036_10_06.bmp, -1, -1, -1
+A0036_10_07.bmp, -1, -1, -1
+A0036_10_08.bmp, -1, -1, -1
+A0036_10_10.bmp, -1, -1, -1
+A0036_10_11.bmp, -1, -1, -1
+A0036_10_12.bmp, -1, -1, -1
+A0036_10_14.bmp, -1, -1, -1
+A0036_10_15.bmp, -1, -1, -1
+A0036_10_16.bmp, -1, -1, -1
+A0036_10_17.bmp, -1, -1, -1
+A0036_10_18.bmp, -1, -1, -1
+A0036_10_19.bmp, -1, -1, -1
+A0036_10_20.bmp, -1, -1, -1
+A0036_10_21.bmp, -1, -1, -1
+A0036_10_24.bmp, -1, -1, -1
+A0036_10_26.bmp, -1, -1, -1
+A0036_10_27.bmp, -1, -1, -1
+A0036_10_28.bmp, -1, -1, -1
+A0036_10_29.bmp, -1, -1, -1
+A0036_10_30.bmp, -1, -1, -1
+A0036_10_31.bmp, -1, -1, -1
+A0036_10_32.bmp, -1, -1, -1
+A0036_10_34.bmp, -1, -1, -1
+A0036_10_35.bmp, -1, -1, -1
+A0036_10_36.bmp, -1, -1, -1
+A0036_10_37.bmp, -1, -1, -1
+A0036_10_38.bmp, -1, -1, -1
+A0036_10_39.bmp, -1, -1, -1
+A0036_10_40.bmp, -1, -1, -1
+A0036_10_41.bmp, -1, -1, -1
+A0036_10_42.bmp, -1, -1, -1
+A0036_10_43.bmp, -1, -1, -1
+A0036_10_44.bmp, -1, -1, -1
+A0036_10_46.bmp, -1, -1, -1
+A0036_10_47.bmp, -1, -1, -1
+A0036_10_48.bmp, -1, -1, -1
+A0036_10_49.bmp, -1, -1, -1
+A0036_10_50.bmp, -1, -1, -1
+A0036_10_51.bmp, -1, -1, -1
+A0036_10_52.bmp, -1, -1, -1
+A0036_10_53.bmp, -1, -1, -1
+A0036_10_54.bmp, -1, -1, -1
+A0036_10_55.bmp, -1, -1, -1
+A0036_10_56.bmp, -1, -1, -1
+A0036_10_58.bmp, -1, -1, -1
+A0036_10_61.bmp, -1, -1, -1
+A0036_10_62.bmp, -1, -1, -1
+A0036_10_63.bmp, -1, -1, -1
+A0036_10_64.bmp, -1, -1, -1
+A0036_10_65.bmp, -1, -1, -1
+A0036_10_66.bmp, -1, -1, -1
+A0036_10_67.bmp, -1, -1, -1
+A0036_10_68.bmp, -1, -1, -1
+A0036_10_69.bmp, -1, -1, -1
+A0036_10_70.bmp, -1, -1, -1
+A0036_10_71.bmp, -1, -1, -1
+A0036_10_72.bmp, -1, -1, -1
+A0036_10_73.bmp, -1, -1, -1
+A0036_10_74.bmp, -1, -1, -1
+A0036_10_75.bmp, -1, -1, -1
+A0071_10_00.bmp, -1, -1, -1
+A0071_10_01.bmp, -1, -1, -1
+A0071_10_02.bmp, -1, -1, -1
+A0071_10_03.bmp, -1, -1, -1
+A0071_10_04.bmp, -1, -1, -1
+A0071_10_05.bmp, -1, -1, -1
+A0071_10_06.bmp, -1, -1, -1
+A0071_10_07.bmp, -1, -1, -1
+A0071_10_08.bmp, -1, -1, -1
+A0071_10_10.bmp, -1, -1, -1
+A0071_10_11.bmp, -1, -1, -1
+A0071_10_12.bmp, -1, -1, -1
+A0071_10_14.bmp, -1, -1, -1
+A0071_10_15.bmp, -1, -1, -1
+A0071_10_16.bmp, -1, -1, -1
+A0071_10_17.bmp, -1, -1, -1
+A0071_10_18.bmp, -1, -1, -1
+A0071_10_19.bmp, -1, -1, -1
+A0071_10_20.bmp, -1, -1, -1
+A0071_10_21.bmp, -1, -1, -1
+A0071_10_24.bmp, -1, -1, -1
+A0071_10_26.bmp, -1, -1, -1
+A0071_10_27.bmp, -1, -1, -1
+A0071_10_28.bmp, -1, -1, -1
+A0071_10_29.bmp, -1, -1, -1
+A0071_10_30.bmp, -1, -1, -1
+A0071_10_31.bmp, -1, -1, -1
+A0071_10_32.bmp, -1, -1, -1
+A0071_10_34.bmp, -1, -1, -1
+A0071_10_35.bmp, -1, -1, -1
+A0071_10_36.bmp, -1, -1, -1
+A0071_10_37.bmp, -1, -1, -1
+A0071_10_38.bmp, -1, -1, -1
+A0071_10_39.bmp, -1, -1, -1
+A0071_10_40.bmp, -1, -1, -1
+A0071_10_41.bmp, -1, -1, -1
+A0071_10_42.bmp, -1, -1, -1
+A0071_10_43.bmp, -1, -1, -1
+A0071_10_44.bmp, -1, -1, -1
+A0071_10_46.bmp, -1, -1, -1
+A0071_10_47.bmp, -1, -1, -1
+A0071_10_48.bmp, -1, -1, -1
+A0071_10_49.bmp, -1, -1, -1
+A0071_10_50.bmp, -1, -1, -1
+A0071_10_51.bmp, -1, -1, -1
+A0071_10_52.bmp, -1, -1, -1
+A0071_10_53.bmp, -1, -1, -1
+A0071_10_54.bmp, -1, -1, -1
+A0071_10_55.bmp, -1, -1, -1
+A0071_10_56.bmp, -1, -1, -1
+A0071_10_58.bmp, -1, -1, -1
+A0071_10_61.bmp, -1, -1, -1
+A0071_10_62.bmp, -1, -1, -1
+A0071_10_63.bmp, -1, -1, -1
+A0071_10_64.bmp, -1, -1, -1
+A0071_10_65.bmp, -1, -1, -1
+A0071_10_66.bmp, -1, -1, -1
+A0071_10_67.bmp, -1, -1, -1
+A0071_10_68.bmp, -1, -1, -1
+A0071_10_69.bmp, -1, -1, -1
+A0071_10_70.bmp, -1, -1, -1
+A0071_10_71.bmp, -1, -1, -1
+A0071_10_72.bmp, -1, -1, -1
+A0071_10_73.bmp, -1, -1, -1
+A0071_10_74.bmp, -1, -1, -1
+A0071_10_75.bmp, -1, -1, -1
+A0089_10_00.bmp, -1, -1, -1
+A0089_10_01.bmp, -1, -1, -1
+A0089_10_02.bmp, -1, -1, -1
+A0089_10_03.bmp, -1, -1, -1
+A0089_10_04.bmp, -1, -1, -1
+A0089_10_05.bmp, -1, -1, -1
+A0089_10_06.bmp, -1, -1, -1
+A0089_10_07.bmp, -1, -1, -1
+A0089_10_08.bmp, -1, -1, -1
+A0089_10_10.bmp, -1, -1, -1
+A0089_10_11.bmp, -1, -1, -1
+A0089_10_12.bmp, -1, -1, -1
+A0089_10_14.bmp, -1, -1, -1
+A0089_10_15.bmp, -1, -1, -1
+A0089_10_16.bmp, -1, -1, -1
+A0089_10_17.bmp, -1, -1, -1
+A0089_10_18.bmp, -1, -1, -1
+A0089_10_19.bmp, -1, -1, -1
+A0089_10_20.bmp, -1, -1, -1
+A0089_10_21.bmp, -1, -1, -1
+A0089_10_24.bmp, -1, -1, -1
+A0089_10_26.bmp, -1, -1, -1
+A0089_10_27.bmp, -1, -1, -1
+A0089_10_28.bmp, -1, -1, -1
+A0089_10_29.bmp, -1, -1, -1
+A0089_10_30.bmp, -1, -1, -1
+A0089_10_31.bmp, -1, -1, -1
+A0089_10_32.bmp, -1, -1, -1
+A0089_10_34.bmp, -1, -1, -1
+A0089_10_35.bmp, -1, -1, -1
+A0089_10_36.bmp, -1, -1, -1
+A0089_10_37.bmp, -1, -1, -1
+A0089_10_38.bmp, -1, -1, -1
+A0089_10_39.bmp, -1, -1, -1
+A0089_10_40.bmp, -1, -1, -1
+A0089_10_41.bmp, -1, -1, -1
+A0089_10_42.bmp, -1, -1, -1
+A0089_10_43.bmp, -1, -1, -1
+A0089_10_44.bmp, -1, -1, -1
+A0089_10_46.bmp, -1, -1, -1
+A0089_10_47.bmp, -1, -1, -1
+A0089_10_48.bmp, -1, -1, -1
+A0089_10_49.bmp, -1, -1, -1
+A0089_10_50.bmp, -1, -1, -1
+A0089_10_51.bmp, -1, -1, -1
+A0089_10_52.bmp, -1, -1, -1
+A0089_10_53.bmp, -1, -1, -1
+A0089_10_54.bmp, -1, -1, -1
+A0089_10_55.bmp, -1, -1, -1
+A0089_10_56.bmp, -1, -1, -1
+A0089_10_58.bmp, -1, -1, -1
+A0089_10_61.bmp, -1, -1, -1
+A0089_10_62.bmp, -1, -1, -1
+A0089_10_63.bmp, -1, -1, -1
+A0089_10_64.bmp, -1, -1, -1
+A0089_10_65.bmp, -1, -1, -1
+A0089_10_66.bmp, -1, -1, -1
+A0089_10_67.bmp, -1, -1, -1
+A0089_10_68.bmp, -1, -1, -1
+A0089_10_69.bmp, -1, -1, -1
+A0089_10_70.bmp, -1, -1, -1
+A0089_10_71.bmp, -1, -1, -1
+A0089_10_72.bmp, -1, -1, -1
+A0089_10_73.bmp, -1, -1, -1
+A0089_10_74.bmp, -1, -1, -1
+A0089_10_75.bmp, -1, -1, -1
+A0095_10_00.bmp, -1, -1, -1
+A0095_10_01.bmp, -1, -1, -1
+A0095_10_02.bmp, -1, -1, -1
+A0095_10_03.bmp, -1, -1, -1
+A0095_10_04.bmp, -1, -1, -1
+A0095_10_05.bmp, -1, -1, -1
+A0095_10_06.bmp, -1, -1, -1
+A0095_10_07.bmp, -1, -1, -1
+A0095_10_08.bmp, -1, -1, -1
+A0095_10_10.bmp, -1, -1, -1
+A0095_10_11.bmp, -1, -1, -1
+A0095_10_12.bmp, -1, -1, -1
+A0095_10_14.bmp, -1, -1, -1
+A0095_10_15.bmp, -1, -1, -1
+A0095_10_16.bmp, -1, -1, -1
+A0095_10_17.bmp, -1, -1, -1
+A0095_10_18.bmp, -1, -1, -1
+A0095_10_19.bmp, -1, -1, -1
+A0095_10_20.bmp, -1, -1, -1
+A0095_10_21.bmp, -1, -1, -1
+A0095_10_24.bmp, -1, -1, -1
+A0095_10_26.bmp, -1, -1, -1
+A0095_10_27.bmp, -1, -1, -1
+A0095_10_28.bmp, -1, -1, -1
+A0095_10_29.bmp, -1, -1, -1
+A0095_10_30.bmp, -1, -1, -1
+A0095_10_31.bmp, -1, -1, -1
+A0095_10_32.bmp, -1, -1, -1
+A0095_10_34.bmp, -1, -1, -1
+A0095_10_35.bmp, -1, -1, -1
+A0095_10_36.bmp, -1, -1, -1
+A0095_10_37.bmp, -1, -1, -1
+A0095_10_38.bmp, -1, -1, -1
+A0095_10_39.bmp, -1, -1, -1
+A0095_10_40.bmp, -1, -1, -1
+A0095_10_41.bmp, -1, -1, -1
+A0095_10_42.bmp, -1, -1, -1
+A0095_10_43.bmp, -1, -1, -1
+A0095_10_44.bmp, -1, -1, -1
+A0095_10_46.bmp, -1, -1, -1
+A0095_10_47.bmp, -1, -1, -1
+A0095_10_48.bmp, -1, -1, -1
+A0095_10_49.bmp, -1, -1, -1
+A0095_10_50.bmp, -1, -1, -1
+A0095_10_51.bmp, -1, -1, -1
+A0095_10_52.bmp, -1, -1, -1
+A0095_10_53.bmp, -1, -1, -1
+A0095_10_54.bmp, -1, -1, -1
+A0095_10_55.bmp, -1, -1, -1
+A0095_10_56.bmp, -1, -1, -1
+A0095_10_58.bmp, -1, -1, -1
+A0095_10_61.bmp, -1, -1, -1
+A0095_10_62.bmp, -1, -1, -1
+A0095_10_63.bmp, -1, -1, -1
+A0095_10_64.bmp, -1, -1, -1
+A0095_10_65.bmp, -1, -1, -1
+A0095_10_66.bmp, -1, -1, -1
+A0095_10_67.bmp, -1, -1, -1
+A0095_10_68.bmp, -1, -1, -1
+A0095_10_69.bmp, -1, -1, -1
+A0095_10_70.bmp, -1, -1, -1
+A0095_10_71.bmp, -1, -1, -1
+A0095_10_72.bmp, -1, -1, -1
+A0095_10_73.bmp, -1, -1, -1
+A0095_10_74.bmp, -1, -1, -1
+A0095_10_75.bmp, -1, -1, -1
+A0111_10_00.bmp, -1, -1, -1
+A0111_10_01.bmp, -1, -1, -1
+A0111_10_02.bmp, -1, -1, -1
+A0111_10_03.bmp, -1, -1, -1
+A0111_10_04.bmp, -1, -1, -1
+A0111_10_05.bmp, -1, -1, -1
+A0111_10_06.bmp, -1, -1, -1
+A0111_10_07.bmp, -1, -1, -1
+A0111_10_08.bmp, -1, -1, -1
+A0111_10_10.bmp, -1, -1, -1
+A0111_10_11.bmp, -1, -1, -1
+A0111_10_12.bmp, -1, -1, -1
+A0111_10_14.bmp, -1, -1, -1
+A0111_10_15.bmp, -1, -1, -1
+A0111_10_16.bmp, -1, -1, -1
+A0111_10_17.bmp, -1, -1, -1
+A0111_10_18.bmp, -1, -1, -1
+A0111_10_19.bmp, -1, -1, -1
+A0111_10_20.bmp, -1, -1, -1
+A0111_10_21.bmp, -1, -1, -1
+A0111_10_24.bmp, -1, -1, -1
+A0111_10_26.bmp, -1, -1, -1
+A0111_10_27.bmp, -1, -1, -1
+A0111_10_28.bmp, -1, -1, -1
+A0111_10_29.bmp, -1, -1, -1
+A0111_10_30.bmp, -1, -1, -1
+A0111_10_31.bmp, -1, -1, -1
+A0111_10_32.bmp, -1, -1, -1
+A0111_10_34.bmp, -1, -1, -1
+A0111_10_35.bmp, -1, -1, -1
+A0111_10_36.bmp, -1, -1, -1
+A0111_10_37.bmp, -1, -1, -1
+A0111_10_38.bmp, -1, -1, -1
+A0111_10_39.bmp, -1, -1, -1
+A0111_10_40.bmp, -1, -1, -1
+A0111_10_41.bmp, -1, -1, -1
+A0111_10_42.bmp, -1, -1, -1
+A0111_10_43.bmp, -1, -1, -1
+A0111_10_44.bmp, -1, -1, -1
+A0111_10_46.bmp, -1, -1, -1
+A0111_10_47.bmp, -1, -1, -1
+A0111_10_48.bmp, -1, -1, -1
+A0111_10_49.bmp, -1, -1, -1
+A0111_10_50.bmp, -1, -1, -1
+A0111_10_51.bmp, -1, -1, -1
+A0111_10_52.bmp, -1, -1, -1
+A0111_10_53.bmp, -1, -1, -1
+A0111_10_54.bmp, -1, -1, -1
+A0111_10_55.bmp, -1, -1, -1
+A0111_10_56.bmp, -1, -1, -1
+A0111_10_58.bmp, -1, -1, -1
+A0111_10_61.bmp, -1, -1, -1
+A0111_10_62.bmp, -1, -1, -1
+A0111_10_63.bmp, -1, -1, -1
+A0111_10_64.bmp, -1, -1, -1
+A0111_10_65.bmp, -1, -1, -1
+A0111_10_66.bmp, -1, -1, -1
+A0111_10_67.bmp, -1, -1, -1
+A0111_10_68.bmp, -1, -1, -1
+A0111_10_69.bmp, -1, -1, -1
+A0111_10_70.bmp, -1, -1, -1
+A0111_10_71.bmp, -1, -1, -1
+A0111_10_72.bmp, -1, -1, -1
+A0111_10_73.bmp, -1, -1, -1
+A0111_10_74.bmp, -1, -1, -1
+A0111_10_75.bmp, -1, -1, -1
+A0122_10_00.bmp, -1, -1, -1
+A0122_10_01.bmp, -1, -1, -1
+A0122_10_02.bmp, -1, -1, -1
+A0122_10_03.bmp, -1, -1, -1
+A0122_10_04.bmp, -1, -1, -1
+A0122_10_05.bmp, -1, -1, -1
+A0122_10_06.bmp, -1, -1, -1
+A0122_10_07.bmp, -1, -1, -1
+A0122_10_08.bmp, -1, -1, -1
+A0122_10_10.bmp, -1, -1, -1
+A0122_10_11.bmp, -1, -1, -1
+A0122_10_12.bmp, -1, -1, -1
+A0122_10_14.bmp, -1, -1, -1
+A0122_10_15.bmp, -1, -1, -1
+A0122_10_16.bmp, -1, -1, -1
+A0122_10_17.bmp, -1, -1, -1
+A0122_10_18.bmp, -1, -1, -1
+A0122_10_19.bmp, -1, -1, -1
+A0122_10_20.bmp, -1, -1, -1
+A0122_10_21.bmp, -1, -1, -1
+A0122_10_24.bmp, -1, -1, -1
+A0122_10_26.bmp, -1, -1, -1
+A0122_10_27.bmp, -1, -1, -1
+A0122_10_28.bmp, -1, -1, -1
+A0122_10_29.bmp, -1, -1, -1
+A0122_10_30.bmp, -1, -1, -1
+A0122_10_31.bmp, -1, -1, -1
+A0122_10_32.bmp, -1, -1, -1
+A0122_10_34.bmp, -1, -1, -1
+A0122_10_35.bmp, -1, -1, -1
+A0122_10_36.bmp, -1, -1, -1
+A0122_10_37.bmp, -1, -1, -1
+A0122_10_38.bmp, -1, -1, -1
+A0122_10_39.bmp, -1, -1, -1
+A0122_10_40.bmp, -1, -1, -1
+A0122_10_41.bmp, -1, -1, -1
+A0122_10_42.bmp, -1, -1, -1
+A0122_10_43.bmp, -1, -1, -1
+A0122_10_44.bmp, -1, -1, -1
+A0122_10_46.bmp, -1, -1, -1
+A0122_10_47.bmp, -1, -1, -1
+A0122_10_48.bmp, -1, -1, -1
+A0122_10_49.bmp, -1, -1, -1
+A0122_10_50.bmp, -1, -1, -1
+A0122_10_51.bmp, -1, -1, -1
+A0122_10_52.bmp, -1, -1, -1
+A0122_10_53.bmp, -1, -1, -1
+A0122_10_54.bmp, -1, -1, -1
+A0122_10_55.bmp, -1, -1, -1
+A0122_10_56.bmp, -1, -1, -1
+A0122_10_58.bmp, -1, -1, -1
+A0122_10_61.bmp, -1, -1, -1
+A0122_10_62.bmp, -1, -1, -1
+A0122_10_63.bmp, -1, -1, -1
+A0122_10_64.bmp, -1, -1, -1
+A0122_10_65.bmp, -1, -1, -1
+A0122_10_66.bmp, -1, -1, -1
+A0122_10_67.bmp, -1, -1, -1
+A0122_10_68.bmp, -1, -1, -1
+A0122_10_69.bmp, -1, -1, -1
+A0122_10_70.bmp, -1, -1, -1
+A0122_10_71.bmp, -1, -1, -1
+A0122_10_72.bmp, -1, -1, -1
+A0122_10_73.bmp, -1, -1, -1
+A0122_10_74.bmp, -1, -1, -1
+A0122_10_75.bmp, -1, -1, -1
+A0126_10_00.bmp, -1, -1, -1
+A0126_10_01.bmp, -1, -1, -1
+A0126_10_02.bmp, -1, -1, -1
+A0126_10_03.bmp, -1, -1, -1
+A0126_10_04.bmp, -1, -1, -1
+A0126_10_05.bmp, -1, -1, -1
+A0126_10_06.bmp, -1, -1, -1
+A0126_10_07.bmp, -1, -1, -1
+A0126_10_08.bmp, -1, -1, -1
+A0126_10_10.bmp, -1, -1, -1
+A0126_10_11.bmp, -1, -1, -1
+A0126_10_12.bmp, -1, -1, -1
+A0126_10_14.bmp, -1, -1, -1
+A0126_10_15.bmp, -1, -1, -1
+A0126_10_16.bmp, -1, -1, -1
+A0126_10_17.bmp, -1, -1, -1
+A0126_10_18.bmp, -1, -1, -1
+A0126_10_19.bmp, -1, -1, -1
+A0126_10_20.bmp, -1, -1, -1
+A0126_10_21.bmp, -1, -1, -1
+A0126_10_24.bmp, -1, -1, -1
+A0126_10_26.bmp, -1, -1, -1
+A0126_10_27.bmp, -1, -1, -1
+A0126_10_28.bmp, -1, -1, -1
+A0126_10_29.bmp, -1, -1, -1
+A0126_10_30.bmp, -1, -1, -1
+A0126_10_31.bmp, -1, -1, -1
+A0126_10_32.bmp, -1, -1, -1
+A0126_10_34.bmp, -1, -1, -1
+A0126_10_35.bmp, -1, -1, -1
+A0126_10_36.bmp, -1, -1, -1
+A0126_10_37.bmp, -1, -1, -1
+A0126_10_38.bmp, -1, -1, -1
+A0126_10_39.bmp, -1, -1, -1
+A0126_10_40.bmp, -1, -1, -1
+A0126_10_41.bmp, -1, -1, -1
+A0126_10_42.bmp, -1, -1, -1
+A0126_10_43.bmp, -1, -1, -1
+A0126_10_44.bmp, -1, -1, -1
+A0126_10_46.bmp, -1, -1, -1
+A0126_10_47.bmp, -1, -1, -1
+A0126_10_48.bmp, -1, -1, -1
+A0126_10_49.bmp, -1, -1, -1
+A0126_10_50.bmp, -1, -1, -1
+A0126_10_51.bmp, -1, -1, -1
+A0126_10_52.bmp, -1, -1, -1
+A0126_10_53.bmp, -1, -1, -1
+A0126_10_54.bmp, -1, -1, -1
+A0126_10_55.bmp, -1, -1, -1
+A0126_10_56.bmp, -1, -1, -1
+A0126_10_58.bmp, -1, -1, -1
+A0126_10_61.bmp, -1, -1, -1
+A0126_10_62.bmp, -1, -1, -1
+A0126_10_63.bmp, -1, -1, -1
+A0126_10_64.bmp, -1, -1, -1
+A0126_10_65.bmp, -1, -1, -1
+A0126_10_66.bmp, -1, -1, -1
+A0126_10_67.bmp, -1, -1, -1
+A0126_10_68.bmp, -1, -1, -1
+A0126_10_69.bmp, -1, -1, -1
+A0126_10_70.bmp, -1, -1, -1
+A0126_10_71.bmp, -1, -1, -1
+A0126_10_72.bmp, -1, -1, -1
+A0126_10_73.bmp, -1, -1, -1
+A0126_10_74.bmp, -1, -1, -1
+A0126_10_75.bmp, -1, -1, -1
+A0127_10_00.bmp, -1, -1, -1
+A0127_10_01.bmp, -1, -1, -1
+A0127_10_02.bmp, -1, -1, -1
+A0127_10_03.bmp, -1, -1, -1
+A0127_10_04.bmp, -1, -1, -1
+A0127_10_05.bmp, -1, -1, -1
+A0127_10_06.bmp, -1, -1, -1
+A0127_10_07.bmp, -1, -1, -1
+A0127_10_08.bmp, -1, -1, -1
+A0127_10_10.bmp, -1, -1, -1
+A0127_10_11.bmp, -1, -1, -1
+A0127_10_12.bmp, -1, -1, -1
+A0127_10_14.bmp, -1, -1, -1
+A0127_10_15.bmp, -1, -1, -1
+A0127_10_16.bmp, -1, -1, -1
+A0127_10_17.bmp, -1, -1, -1
+A0127_10_18.bmp, -1, -1, -1
+A0127_10_19.bmp, -1, -1, -1
+A0127_10_20.bmp, -1, -1, -1
+A0127_10_21.bmp, -1, -1, -1
+A0127_10_24.bmp, -1, -1, -1
+A0127_10_26.bmp, -1, -1, -1
+A0127_10_27.bmp, -1, -1, -1
+A0127_10_28.bmp, -1, -1, -1
+A0127_10_29.bmp, -1, -1, -1
+A0127_10_30.bmp, -1, -1, -1
+A0127_10_31.bmp, -1, -1, -1
+A0127_10_32.bmp, -1, -1, -1
+A0127_10_34.bmp, -1, -1, -1
+A0127_10_35.bmp, -1, -1, -1
+A0127_10_36.bmp, -1, -1, -1
+A0127_10_37.bmp, -1, -1, -1
+A0127_10_38.bmp, -1, -1, -1
+A0127_10_39.bmp, -1, -1, -1
+A0127_10_40.bmp, -1, -1, -1
+A0127_10_41.bmp, -1, -1, -1
+A0127_10_42.bmp, -1, -1, -1
+A0127_10_43.bmp, -1, -1, -1
+A0127_10_44.bmp, -1, -1, -1
+A0127_10_46.bmp, -1, -1, -1
+A0127_10_47.bmp, -1, -1, -1
+A0127_10_48.bmp, -1, -1, -1
+A0127_10_49.bmp, -1, -1, -1
+A0127_10_50.bmp, -1, -1, -1
+A0127_10_51.bmp, -1, -1, -1
+A0127_10_52.bmp, -1, -1, -1
+A0127_10_53.bmp, -1, -1, -1
+A0127_10_54.bmp, -1, -1, -1
+A0127_10_55.bmp, -1, -1, -1
+A0127_10_56.bmp, -1, -1, -1
+A0127_10_58.bmp, -1, -1, -1
+A0127_10_61.bmp, -1, -1, -1
+A0127_10_62.bmp, -1, -1, -1
+A0127_10_63.bmp, -1, -1, -1
+A0127_10_64.bmp, -1, -1, -1
+A0127_10_65.bmp, -1, -1, -1
+A0127_10_66.bmp, -1, -1, -1
+A0127_10_67.bmp, -1, -1, -1
+A0127_10_68.bmp, -1, -1, -1
+A0127_10_69.bmp, -1, -1, -1
+A0127_10_70.bmp, -1, -1, -1
+A0127_10_71.bmp, -1, -1, -1
+A0127_10_72.bmp, -1, -1, -1
+A0127_10_73.bmp, -1, -1, -1
+A0127_10_74.bmp, -1, -1, -1
+A0127_10_75.bmp, -1, -1, -1
+A0129_10_00.bmp, -1, -1, -1
+A0129_10_01.bmp, -1, -1, -1
+A0129_10_02.bmp, -1, -1, -1
+A0129_10_03.bmp, -1, -1, -1
+A0129_10_04.bmp, -1, -1, -1
+A0129_10_05.bmp, -1, -1, -1
+A0129_10_06.bmp, -1, -1, -1
+A0129_10_07.bmp, -1, -1, -1
+A0129_10_08.bmp, -1, -1, -1
+A0129_10_10.bmp, -1, -1, -1
+A0129_10_11.bmp, -1, -1, -1
+A0129_10_12.bmp, -1, -1, -1
+A0129_10_14.bmp, -1, -1, -1
+A0129_10_15.bmp, -1, -1, -1
+A0129_10_16.bmp, -1, -1, -1
+A0129_10_17.bmp, -1, -1, -1
+A0129_10_18.bmp, -1, -1, -1
+A0129_10_19.bmp, -1, -1, -1
+A0129_10_20.bmp, -1, -1, -1
+A0129_10_21.bmp, -1, -1, -1
+A0129_10_24.bmp, -1, -1, -1
+A0129_10_26.bmp, -1, -1, -1
+A0129_10_27.bmp, -1, -1, -1
+A0129_10_28.bmp, -1, -1, -1
+A0129_10_29.bmp, -1, -1, -1
+A0129_10_30.bmp, -1, -1, -1
+A0129_10_31.bmp, -1, -1, -1
+A0129_10_32.bmp, -1, -1, -1
+A0129_10_34.bmp, -1, -1, -1
+A0129_10_35.bmp, -1, -1, -1
+A0129_10_36.bmp, -1, -1, -1
+A0129_10_37.bmp, -1, -1, -1
+A0129_10_38.bmp, -1, -1, -1
+A0129_10_39.bmp, -1, -1, -1
+A0129_10_40.bmp, -1, -1, -1
+A0129_10_41.bmp, -1, -1, -1
+A0129_10_42.bmp, -1, -1, -1
+A0129_10_43.bmp, -1, -1, -1
+A0129_10_44.bmp, -1, -1, -1
+A0129_10_46.bmp, -1, -1, -1
+A0129_10_47.bmp, -1, -1, -1
+A0129_10_48.bmp, -1, -1, -1
+A0129_10_49.bmp, -1, -1, -1
+A0129_10_50.bmp, -1, -1, -1
+A0129_10_51.bmp, -1, -1, -1
+A0129_10_52.bmp, -1, -1, -1
+A0129_10_53.bmp, -1, -1, -1
+A0129_10_54.bmp, -1, -1, -1
+A0129_10_55.bmp, -1, -1, -1
+A0129_10_56.bmp, -1, -1, -1
+A0129_10_58.bmp, -1, -1, -1
+A0129_10_61.bmp, -1, -1, -1
+A0129_10_62.bmp, -1, -1, -1
+A0129_10_63.bmp, -1, -1, -1
+A0129_10_64.bmp, -1, -1, -1
+A0129_10_65.bmp, -1, -1, -1
+A0129_10_66.bmp, -1, -1, -1
+A0129_10_67.bmp, -1, -1, -1
+A0129_10_68.bmp, -1, -1, -1
+A0129_10_69.bmp, -1, -1, -1
+A0129_10_70.bmp, -1, -1, -1
+A0129_10_71.bmp, -1, -1, -1
+A0129_10_72.bmp, -1, -1, -1
+A0129_10_73.bmp, -1, -1, -1
+A0129_10_74.bmp, -1, -1, -1
+A0129_10_75.bmp, -1, -1, -1
+A0144_10_00.bmp, -1, -1, -1
+A0144_10_01.bmp, -1, -1, -1
+A0144_10_02.bmp, -1, -1, -1
+A0144_10_03.bmp, -1, -1, -1
+A0144_10_04.bmp, -1, -1, -1
+A0144_10_05.bmp, -1, -1, -1
+A0144_10_06.bmp, -1, -1, -1
+A0144_10_07.bmp, -1, -1, -1
+A0144_10_08.bmp, -1, -1, -1
+A0144_10_10.bmp, -1, -1, -1
+A0144_10_11.bmp, -1, -1, -1
+A0144_10_12.bmp, -1, -1, -1
+A0144_10_14.bmp, -1, -1, -1
+A0144_10_15.bmp, -1, -1, -1
+A0144_10_16.bmp, -1, -1, -1
+A0144_10_17.bmp, -1, -1, -1
+A0144_10_18.bmp, -1, -1, -1
+A0144_10_19.bmp, -1, -1, -1
+A0144_10_20.bmp, -1, -1, -1
+A0144_10_21.bmp, -1, -1, -1
+A0144_10_24.bmp, -1, -1, -1
+A0144_10_26.bmp, -1, -1, -1
+A0144_10_27.bmp, -1, -1, -1
+A0144_10_28.bmp, -1, -1, -1
+A0144_10_29.bmp, -1, -1, -1
+A0144_10_30.bmp, -1, -1, -1
+A0144_10_31.bmp, -1, -1, -1
+A0144_10_32.bmp, -1, -1, -1
+A0144_10_34.bmp, -1, -1, -1
+A0144_10_35.bmp, -1, -1, -1
+A0144_10_36.bmp, -1, -1, -1
+A0144_10_37.bmp, -1, -1, -1
+A0144_10_38.bmp, -1, -1, -1
+A0144_10_39.bmp, -1, -1, -1
+A0144_10_40.bmp, -1, -1, -1
+A0144_10_41.bmp, -1, -1, -1
+A0144_10_42.bmp, -1, -1, -1
+A0144_10_43.bmp, -1, -1, -1
+A0144_10_44.bmp, -1, -1, -1
+A0144_10_46.bmp, -1, -1, -1
+A0144_10_47.bmp, -1, -1, -1
+A0144_10_48.bmp, -1, -1, -1
+A0144_10_49.bmp, -1, -1, -1
+A0144_10_50.bmp, -1, -1, -1
+A0144_10_51.bmp, -1, -1, -1
+A0144_10_52.bmp, -1, -1, -1
+A0144_10_53.bmp, -1, -1, -1
+A0144_10_54.bmp, -1, -1, -1
+A0144_10_55.bmp, -1, -1, -1
+A0144_10_56.bmp, -1, -1, -1
+A0144_10_58.bmp, -1, -1, -1
+A0144_10_61.bmp, -1, -1, -1
+A0144_10_62.bmp, -1, -1, -1
+A0144_10_63.bmp, -1, -1, -1
+A0144_10_64.bmp, -1, -1, -1
+A0144_10_65.bmp, -1, -1, -1
+A0144_10_66.bmp, -1, -1, -1
+A0144_10_67.bmp, -1, -1, -1
+A0144_10_68.bmp, -1, -1, -1
+A0144_10_69.bmp, -1, -1, -1
+A0144_10_70.bmp, -1, -1, -1
+A0144_10_71.bmp, -1, -1, -1
+A0144_10_72.bmp, -1, -1, -1
+A0144_10_73.bmp, -1, -1, -1
+A0144_10_74.bmp, -1, -1, -1
+A0144_10_75.bmp, -1, -1, -1
+A0147_10_00.bmp, -1, -1, -1
+A0147_10_01.bmp, -1, -1, -1
+A0147_10_02.bmp, -1, -1, -1
+A0147_10_03.bmp, -1, -1, -1
+A0147_10_04.bmp, -1, -1, -1
+A0147_10_05.bmp, -1, -1, -1
+A0147_10_06.bmp, -1, -1, -1
+A0147_10_07.bmp, -1, -1, -1
+A0147_10_08.bmp, -1, -1, -1
+A0147_10_10.bmp, -1, -1, -1
+A0147_10_11.bmp, -1, -1, -1
+A0147_10_12.bmp, -1, -1, -1
+A0147_10_14.bmp, -1, -1, -1
+A0147_10_15.bmp, -1, -1, -1
+A0147_10_16.bmp, -1, -1, -1
+A0147_10_17.bmp, -1, -1, -1
+A0147_10_18.bmp, -1, -1, -1
+A0147_10_19.bmp, -1, -1, -1
+A0147_10_20.bmp, -1, -1, -1
+A0147_10_21.bmp, -1, -1, -1
+A0147_10_24.bmp, -1, -1, -1
+A0147_10_26.bmp, -1, -1, -1
+A0147_10_27.bmp, -1, -1, -1
+A0147_10_28.bmp, -1, -1, -1
+A0147_10_29.bmp, -1, -1, -1
+A0147_10_30.bmp, -1, -1, -1
+A0147_10_31.bmp, -1, -1, -1
+A0147_10_32.bmp, -1, -1, -1
+A0147_10_34.bmp, -1, -1, -1
+A0147_10_35.bmp, -1, -1, -1
+A0147_10_36.bmp, -1, -1, -1
+A0147_10_37.bmp, -1, -1, -1
+A0147_10_38.bmp, -1, -1, -1
+A0147_10_39.bmp, -1, -1, -1
+A0147_10_40.bmp, -1, -1, -1
+A0147_10_41.bmp, -1, -1, -1
+A0147_10_42.bmp, -1, -1, -1
+A0147_10_43.bmp, -1, -1, -1
+A0147_10_44.bmp, -1, -1, -1
+A0147_10_46.bmp, -1, -1, -1
+A0147_10_47.bmp, -1, -1, -1
+A0147_10_48.bmp, -1, -1, -1
+A0147_10_49.bmp, -1, -1, -1
+A0147_10_50.bmp, -1, -1, -1
+A0147_10_51.bmp, -1, -1, -1
+A0147_10_52.bmp, -1, -1, -1
+A0147_10_53.bmp, -1, -1, -1
+A0147_10_54.bmp, -1, -1, -1
+A0147_10_55.bmp, -1, -1, -1
+A0147_10_56.bmp, -1, -1, -1
+A0147_10_58.bmp, -1, -1, -1
+A0147_10_61.bmp, -1, -1, -1
+A0147_10_62.bmp, -1, -1, -1
+A0147_10_63.bmp, -1, -1, -1
+A0147_10_64.bmp, -1, -1, -1
+A0147_10_65.bmp, -1, -1, -1
+A0147_10_66.bmp, -1, -1, -1
+A0147_10_67.bmp, -1, -1, -1
+A0147_10_68.bmp, -1, -1, -1
+A0147_10_69.bmp, -1, -1, -1
+A0147_10_70.bmp, -1, -1, -1
+A0147_10_71.bmp, -1, -1, -1
+A0147_10_72.bmp, -1, -1, -1
+A0147_10_73.bmp, -1, -1, -1
+A0147_10_74.bmp, -1, -1, -1
+A0147_10_75.bmp, -1, -1, -1
+A0154_10_00.bmp, -1, -1, -1
+A0154_10_01.bmp, -1, -1, -1
+A0154_10_02.bmp, -1, -1, -1
+A0154_10_03.bmp, -1, -1, -1
+A0154_10_04.bmp, -1, -1, -1
+A0154_10_05.bmp, -1, -1, -1
+A0154_10_06.bmp, -1, -1, -1
+A0154_10_07.bmp, -1, -1, -1
+A0154_10_08.bmp, -1, -1, -1
+A0154_10_10.bmp, -1, -1, -1
+A0154_10_11.bmp, -1, -1, -1
+A0154_10_12.bmp, -1, -1, -1
+A0154_10_14.bmp, -1, -1, -1
+A0154_10_15.bmp, -1, -1, -1
+A0154_10_16.bmp, -1, -1, -1
+A0154_10_17.bmp, -1, -1, -1
+A0154_10_18.bmp, -1, -1, -1
+A0154_10_19.bmp, -1, -1, -1
+A0154_10_20.bmp, -1, -1, -1
+A0154_10_21.bmp, -1, -1, -1
+A0154_10_24.bmp, -1, -1, -1
+A0154_10_26.bmp, -1, -1, -1
+A0154_10_27.bmp, -1, -1, -1
+A0154_10_28.bmp, -1, -1, -1
+A0154_10_29.bmp, -1, -1, -1
+A0154_10_30.bmp, -1, -1, -1
+A0154_10_31.bmp, -1, -1, -1
+A0154_10_32.bmp, -1, -1, -1
+A0154_10_34.bmp, -1, -1, -1
+A0154_10_35.bmp, -1, -1, -1
+A0154_10_36.bmp, -1, -1, -1
+A0154_10_37.bmp, -1, -1, -1
+A0154_10_38.bmp, -1, -1, -1
+A0154_10_39.bmp, -1, -1, -1
+A0154_10_40.bmp, -1, -1, -1
+A0154_10_41.bmp, -1, -1, -1
+A0154_10_42.bmp, -1, -1, -1
+A0154_10_43.bmp, -1, -1, -1
+A0154_10_44.bmp, -1, -1, -1
+A0154_10_46.bmp, -1, -1, -1
+A0154_10_47.bmp, -1, -1, -1
+A0154_10_48.bmp, -1, -1, -1
+A0154_10_49.bmp, -1, -1, -1
+A0154_10_50.bmp, -1, -1, -1
+A0154_10_51.bmp, -1, -1, -1
+A0154_10_52.bmp, -1, -1, -1
+A0154_10_53.bmp, -1, -1, -1
+A0154_10_54.bmp, -1, -1, -1
+A0154_10_55.bmp, -1, -1, -1
+A0154_10_56.bmp, -1, -1, -1
+A0154_10_58.bmp, -1, -1, -1
+A0154_10_61.bmp, -1, -1, -1
+A0154_10_62.bmp, -1, -1, -1
+A0154_10_63.bmp, -1, -1, -1
+A0154_10_64.bmp, -1, -1, -1
+A0154_10_65.bmp, -1, -1, -1
+A0154_10_66.bmp, -1, -1, -1
+A0154_10_67.bmp, -1, -1, -1
+A0154_10_68.bmp, -1, -1, -1
+A0154_10_69.bmp, -1, -1, -1
+A0154_10_70.bmp, -1, -1, -1
+A0154_10_71.bmp, -1, -1, -1
+A0154_10_72.bmp, -1, -1, -1
+A0154_10_73.bmp, -1, -1, -1
+A0154_10_74.bmp, -1, -1, -1
+A0154_10_75.bmp, -1, -1, -1
+A0161_10_00.bmp, -1, -1, -1
+A0161_10_01.bmp, -1, -1, -1
+A0161_10_02.bmp, -1, -1, -1
+A0161_10_03.bmp, -1, -1, -1
+A0161_10_04.bmp, -1, -1, -1
+A0161_10_05.bmp, -1, -1, -1
+A0161_10_06.bmp, -1, -1, -1
+A0161_10_07.bmp, -1, -1, -1
+A0161_10_08.bmp, -1, -1, -1
+A0161_10_10.bmp, -1, -1, -1
+A0161_10_11.bmp, -1, -1, -1
+A0161_10_12.bmp, -1, -1, -1
+A0161_10_14.bmp, -1, -1, -1
+A0161_10_15.bmp, -1, -1, -1
+A0161_10_16.bmp, -1, -1, -1
+A0161_10_17.bmp, -1, -1, -1
+A0161_10_18.bmp, -1, -1, -1
+A0161_10_19.bmp, -1, -1, -1
+A0161_10_20.bmp, -1, -1, -1
+A0161_10_21.bmp, -1, -1, -1
+A0161_10_24.bmp, -1, -1, -1
+A0161_10_26.bmp, -1, -1, -1
+A0161_10_27.bmp, -1, -1, -1
+A0161_10_28.bmp, -1, -1, -1
+A0161_10_29.bmp, -1, -1, -1
+A0161_10_30.bmp, -1, -1, -1
+A0161_10_31.bmp, -1, -1, -1
+A0161_10_32.bmp, -1, -1, -1
+A0161_10_34.bmp, -1, -1, -1
+A0161_10_35.bmp, -1, -1, -1
+A0161_10_36.bmp, -1, -1, -1
+A0161_10_37.bmp, -1, -1, -1
+A0161_10_38.bmp, -1, -1, -1
+A0161_10_39.bmp, -1, -1, -1
+A0161_10_40.bmp, -1, -1, -1
+A0161_10_41.bmp, -1, -1, -1
+A0161_10_42.bmp, -1, -1, -1
+A0161_10_43.bmp, -1, -1, -1
+A0161_10_44.bmp, -1, -1, -1
+A0161_10_46.bmp, -1, -1, -1
+A0161_10_47.bmp, -1, -1, -1
+A0161_10_48.bmp, -1, -1, -1
+A0161_10_49.bmp, -1, -1, -1
+A0161_10_50.bmp, -1, -1, -1
+A0161_10_51.bmp, -1, -1, -1
+A0161_10_52.bmp, -1, -1, -1
+A0161_10_53.bmp, -1, -1, -1
+A0161_10_54.bmp, -1, -1, -1
+A0161_10_55.bmp, -1, -1, -1
+A0161_10_56.bmp, -1, -1, -1
+A0161_10_58.bmp, -1, -1, -1
+A0161_10_61.bmp, -1, -1, -1
+A0161_10_62.bmp, -1, -1, -1
+A0161_10_63.bmp, -1, -1, -1
+A0161_10_64.bmp, -1, -1, -1
+A0161_10_65.bmp, -1, -1, -1
+A0161_10_66.bmp, -1, -1, -1
+A0161_10_67.bmp, -1, -1, -1
+A0161_10_68.bmp, -1, -1, -1
+A0161_10_69.bmp, -1, -1, -1
+A0161_10_70.bmp, -1, -1, -1
+A0161_10_71.bmp, -1, -1, -1
+A0161_10_72.bmp, -1, -1, -1
+A0161_10_73.bmp, -1, -1, -1
+A0161_10_74.bmp, -1, -1, -1
+A0161_10_75.bmp, -1, -1, -1
+A0192_10_00.bmp, -1, -1, -1
+A0192_10_01.bmp, -1, -1, -1
+A0192_10_02.bmp, -1, -1, -1
+A0192_10_03.bmp, -1, -1, -1
+A0192_10_04.bmp, -1, -1, -1
+A0192_10_05.bmp, -1, -1, -1
+A0192_10_06.bmp, -1, -1, -1
+A0192_10_07.bmp, -1, -1, -1
+A0192_10_08.bmp, -1, -1, -1
+A0192_10_10.bmp, -1, -1, -1
+A0192_10_11.bmp, -1, -1, -1
+A0192_10_12.bmp, -1, -1, -1
+A0192_10_14.bmp, -1, -1, -1
+A0192_10_15.bmp, -1, -1, -1
+A0192_10_16.bmp, -1, -1, -1
+A0192_10_17.bmp, -1, -1, -1
+A0192_10_18.bmp, -1, -1, -1
+A0192_10_19.bmp, -1, -1, -1
+A0192_10_20.bmp, -1, -1, -1
+A0192_10_21.bmp, -1, -1, -1
+A0192_10_24.bmp, -1, -1, -1
+A0192_10_26.bmp, -1, -1, -1
+A0192_10_27.bmp, -1, -1, -1
+A0192_10_28.bmp, -1, -1, -1
+A0192_10_29.bmp, -1, -1, -1
+A0192_10_30.bmp, -1, -1, -1
+A0192_10_31.bmp, -1, -1, -1
+A0192_10_32.bmp, -1, -1, -1
+A0192_10_34.bmp, -1, -1, -1
+A0192_10_35.bmp, -1, -1, -1
+A0192_10_36.bmp, -1, -1, -1
+A0192_10_37.bmp, -1, -1, -1
+A0192_10_38.bmp, -1, -1, -1
+A0192_10_39.bmp, -1, -1, -1
+A0192_10_40.bmp, -1, -1, -1
+A0192_10_41.bmp, -1, -1, -1
+A0192_10_42.bmp, -1, -1, -1
+A0192_10_43.bmp, -1, -1, -1
+A0192_10_44.bmp, -1, -1, -1
+A0192_10_46.bmp, -1, -1, -1
+A0192_10_47.bmp, -1, -1, -1
+A0192_10_48.bmp, -1, -1, -1
+A0192_10_49.bmp, -1, -1, -1
+A0192_10_50.bmp, -1, -1, -1
+A0192_10_51.bmp, -1, -1, -1
+A0192_10_52.bmp, -1, -1, -1
+A0192_10_53.bmp, -1, -1, -1
+A0192_10_54.bmp, -1, -1, -1
+A0192_10_55.bmp, -1, -1, -1
+A0192_10_56.bmp, -1, -1, -1
+A0192_10_58.bmp, -1, -1, -1
+A0192_10_61.bmp, -1, -1, -1
+A0192_10_62.bmp, -1, -1, -1
+A0192_10_63.bmp, -1, -1, -1
+A0192_10_64.bmp, -1, -1, -1
+A0192_10_65.bmp, -1, -1, -1
+A0192_10_66.bmp, -1, -1, -1
+A0192_10_67.bmp, -1, -1, -1
+A0192_10_68.bmp, -1, -1, -1
+A0192_10_69.bmp, -1, -1, -1
+A0192_10_70.bmp, -1, -1, -1
+A0192_10_71.bmp, -1, -1, -1
+A0192_10_72.bmp, -1, -1, -1
+A0192_10_73.bmp, -1, -1, -1
+A0192_10_74.bmp, -1, -1, -1
+A0192_10_75.bmp, -1, -1, -1
+A0196_10_00.bmp, -1, -1, -1
+A0196_10_01.bmp, -1, -1, -1
+A0196_10_02.bmp, -1, -1, -1
+A0196_10_03.bmp, -1, -1, -1
+A0196_10_04.bmp, -1, -1, -1
+A0196_10_05.bmp, -1, -1, -1
+A0196_10_06.bmp, -1, -1, -1
+A0196_10_07.bmp, -1, -1, -1
+A0196_10_08.bmp, -1, -1, -1
+A0196_10_10.bmp, -1, -1, -1
+A0196_10_11.bmp, -1, -1, -1
+A0196_10_12.bmp, -1, -1, -1
+A0196_10_14.bmp, -1, -1, -1
+A0196_10_15.bmp, -1, -1, -1
+A0196_10_16.bmp, -1, -1, -1
+A0196_10_17.bmp, -1, -1, -1
+A0196_10_18.bmp, -1, -1, -1
+A0196_10_19.bmp, -1, -1, -1
+A0196_10_20.bmp, -1, -1, -1
+A0196_10_21.bmp, -1, -1, -1
+A0196_10_24.bmp, -1, -1, -1
+A0196_10_26.bmp, -1, -1, -1
+A0196_10_27.bmp, -1, -1, -1
+A0196_10_28.bmp, -1, -1, -1
+A0196_10_29.bmp, -1, -1, -1
+A0196_10_30.bmp, -1, -1, -1
+A0196_10_31.bmp, -1, -1, -1
+A0196_10_32.bmp, -1, -1, -1
+A0196_10_34.bmp, -1, -1, -1
+A0196_10_35.bmp, -1, -1, -1
+A0196_10_36.bmp, -1, -1, -1
+A0196_10_37.bmp, -1, -1, -1
+A0196_10_38.bmp, -1, -1, -1
+A0196_10_39.bmp, -1, -1, -1
+A0196_10_40.bmp, -1, -1, -1
+A0196_10_41.bmp, -1, -1, -1
+A0196_10_42.bmp, -1, -1, -1
+A0196_10_43.bmp, -1, -1, -1
+A0196_10_44.bmp, -1, -1, -1
+A0196_10_46.bmp, -1, -1, -1
+A0196_10_47.bmp, -1, -1, -1
+A0196_10_48.bmp, -1, -1, -1
+A0196_10_49.bmp, -1, -1, -1
+A0196_10_50.bmp, -1, -1, -1
+A0196_10_51.bmp, -1, -1, -1
+A0196_10_52.bmp, -1, -1, -1
+A0196_10_53.bmp, -1, -1, -1
+A0196_10_54.bmp, -1, -1, -1
+A0196_10_55.bmp, -1, -1, -1
+A0196_10_56.bmp, -1, -1, -1
+A0196_10_58.bmp, -1, -1, -1
+A0196_10_61.bmp, -1, -1, -1
+A0196_10_62.bmp, -1, -1, -1
+A0196_10_63.bmp, -1, -1, -1
+A0196_10_64.bmp, -1, -1, -1
+A0196_10_65.bmp, -1, -1, -1
+A0196_10_66.bmp, -1, -1, -1
+A0196_10_67.bmp, -1, -1, -1
+A0196_10_68.bmp, -1, -1, -1
+A0196_10_69.bmp, -1, -1, -1
+A0196_10_70.bmp, -1, -1, -1
+A0196_10_71.bmp, -1, -1, -1
+A0196_10_72.bmp, -1, -1, -1
+A0196_10_73.bmp, -1, -1, -1
+A0196_10_74.bmp, -1, -1, -1
+A0196_10_75.bmp, -1, -1, -1
+A0202_10_00.bmp, -1, -1, -1
+A0202_10_01.bmp, -1, -1, -1
+A0202_10_02.bmp, -1, -1, -1
+A0202_10_03.bmp, -1, -1, -1
+A0202_10_04.bmp, -1, -1, -1
+A0202_10_05.bmp, -1, -1, -1
+A0202_10_06.bmp, -1, -1, -1
+A0202_10_07.bmp, -1, -1, -1
+A0202_10_08.bmp, -1, -1, -1
+A0202_10_10.bmp, -1, -1, -1
+A0202_10_11.bmp, -1, -1, -1
+A0202_10_12.bmp, -1, -1, -1
+A0202_10_14.bmp, -1, -1, -1
+A0202_10_15.bmp, -1, -1, -1
+A0202_10_16.bmp, -1, -1, -1
+A0202_10_17.bmp, -1, -1, -1
+A0202_10_18.bmp, -1, -1, -1
+A0202_10_19.bmp, -1, -1, -1
+A0202_10_20.bmp, -1, -1, -1
+A0202_10_21.bmp, -1, -1, -1
+A0202_10_24.bmp, -1, -1, -1
+A0202_10_26.bmp, -1, -1, -1
+A0202_10_27.bmp, -1, -1, -1
+A0202_10_28.bmp, -1, -1, -1
+A0202_10_29.bmp, -1, -1, -1
+A0202_10_30.bmp, -1, -1, -1
+A0202_10_31.bmp, -1, -1, -1
+A0202_10_32.bmp, -1, -1, -1
+A0202_10_34.bmp, -1, -1, -1
+A0202_10_35.bmp, -1, -1, -1
+A0202_10_36.bmp, -1, -1, -1
+A0202_10_37.bmp, -1, -1, -1
+A0202_10_38.bmp, -1, -1, -1
+A0202_10_39.bmp, -1, -1, -1
+A0202_10_40.bmp, -1, -1, -1
+A0202_10_41.bmp, -1, -1, -1
+A0202_10_42.bmp, -1, -1, -1
+A0202_10_43.bmp, -1, -1, -1
+A0202_10_44.bmp, -1, -1, -1
+A0202_10_46.bmp, -1, -1, -1
+A0202_10_47.bmp, -1, -1, -1
+A0202_10_48.bmp, -1, -1, -1
+A0202_10_49.bmp, -1, -1, -1
+A0202_10_50.bmp, -1, -1, -1
+A0202_10_51.bmp, -1, -1, -1
+A0202_10_52.bmp, -1, -1, -1
+A0202_10_53.bmp, -1, -1, -1
+A0202_10_54.bmp, -1, -1, -1
+A0202_10_55.bmp, -1, -1, -1
+A0202_10_56.bmp, -1, -1, -1
+A0202_10_58.bmp, -1, -1, -1
+A0202_10_61.bmp, -1, -1, -1
+A0202_10_62.bmp, -1, -1, -1
+A0202_10_63.bmp, -1, -1, -1
+A0202_10_64.bmp, -1, -1, -1
+A0202_10_65.bmp, -1, -1, -1
+A0202_10_66.bmp, -1, -1, -1
+A0202_10_67.bmp, -1, -1, -1
+A0202_10_68.bmp, -1, -1, -1
+A0202_10_69.bmp, -1, -1, -1
+A0202_10_70.bmp, -1, -1, -1
+A0202_10_71.bmp, -1, -1, -1
+A0202_10_72.bmp, -1, -1, -1
+A0202_10_73.bmp, -1, -1, -1
+A0202_10_74.bmp, -1, -1, -1
+A0202_10_75.bmp, -1, -1, -1
+A0217_10_00.bmp, -1, -1, -1
+A0217_10_01.bmp, -1, -1, -1
+A0217_10_02.bmp, -1, -1, -1
+A0217_10_03.bmp, -1, -1, -1
+A0217_10_04.bmp, -1, -1, -1
+A0217_10_05.bmp, -1, -1, -1
+A0217_10_06.bmp, -1, -1, -1
+A0217_10_07.bmp, -1, -1, -1
+A0217_10_08.bmp, -1, -1, -1
+A0217_10_10.bmp, -1, -1, -1
+A0217_10_11.bmp, -1, -1, -1
+A0217_10_12.bmp, -1, -1, -1
+A0217_10_14.bmp, -1, -1, -1
+A0217_10_15.bmp, -1, -1, -1
+A0217_10_16.bmp, -1, -1, -1
+A0217_10_17.bmp, -1, -1, -1
+A0217_10_18.bmp, -1, -1, -1
+A0217_10_19.bmp, -1, -1, -1
+A0217_10_20.bmp, -1, -1, -1
+A0217_10_21.bmp, -1, -1, -1
+A0217_10_24.bmp, -1, -1, -1
+A0217_10_26.bmp, -1, -1, -1
+A0217_10_27.bmp, -1, -1, -1
+A0217_10_28.bmp, -1, -1, -1
+A0217_10_29.bmp, -1, -1, -1
+A0217_10_30.bmp, -1, -1, -1
+A0217_10_31.bmp, -1, -1, -1
+A0217_10_32.bmp, -1, -1, -1
+A0217_10_34.bmp, -1, -1, -1
+A0217_10_35.bmp, -1, -1, -1
+A0217_10_36.bmp, -1, -1, -1
+A0217_10_37.bmp, -1, -1, -1
+A0217_10_38.bmp, -1, -1, -1
+A0217_10_39.bmp, -1, -1, -1
+A0217_10_40.bmp, -1, -1, -1
+A0217_10_41.bmp, -1, -1, -1
+A0217_10_42.bmp, -1, -1, -1
+A0217_10_43.bmp, -1, -1, -1
+A0217_10_44.bmp, -1, -1, -1
+A0217_10_46.bmp, -1, -1, -1
+A0217_10_47.bmp, -1, -1, -1
+A0217_10_48.bmp, -1, -1, -1
+A0217_10_49.bmp, -1, -1, -1
+A0217_10_50.bmp, -1, -1, -1
+A0217_10_51.bmp, -1, -1, -1
+A0217_10_52.bmp, -1, -1, -1
+A0217_10_53.bmp, -1, -1, -1
+A0217_10_54.bmp, -1, -1, -1
+A0217_10_55.bmp, -1, -1, -1
+A0217_10_56.bmp, -1, -1, -1
+A0217_10_58.bmp, -1, -1, -1
+A0217_10_61.bmp, -1, -1, -1
+A0217_10_62.bmp, -1, -1, -1
+A0217_10_63.bmp, -1, -1, -1
+A0217_10_64.bmp, -1, -1, -1
+A0217_10_65.bmp, -1, -1, -1
+A0217_10_66.bmp, -1, -1, -1
+A0217_10_67.bmp, -1, -1, -1
+A0217_10_68.bmp, -1, -1, -1
+A0217_10_69.bmp, -1, -1, -1
+A0217_10_70.bmp, -1, -1, -1
+A0217_10_71.bmp, -1, -1, -1
+A0217_10_72.bmp, -1, -1, -1
+A0217_10_73.bmp, -1, -1, -1
+A0217_10_74.bmp, -1, -1, -1
+A0217_10_75.bmp, -1, -1, -1
+A0219_10_00.bmp, -1, -1, -1
+A0219_10_01.bmp, -1, -1, -1
+A0219_10_02.bmp, -1, -1, -1
+A0219_10_03.bmp, -1, -1, -1
+A0219_10_04.bmp, -1, -1, -1
+A0219_10_05.bmp, -1, -1, -1
+A0219_10_06.bmp, -1, -1, -1
+A0219_10_07.bmp, -1, -1, -1
+A0219_10_08.bmp, -1, -1, -1
+A0219_10_10.bmp, -1, -1, -1
+A0219_10_11.bmp, -1, -1, -1
+A0219_10_12.bmp, -1, -1, -1
+A0219_10_14.bmp, -1, -1, -1
+A0219_10_15.bmp, -1, -1, -1
+A0219_10_16.bmp, -1, -1, -1
+A0219_10_17.bmp, -1, -1, -1
+A0219_10_18.bmp, -1, -1, -1
+A0219_10_19.bmp, -1, -1, -1
+A0219_10_20.bmp, -1, -1, -1
+A0219_10_21.bmp, -1, -1, -1
+A0219_10_24.bmp, -1, -1, -1
+A0219_10_26.bmp, -1, -1, -1
+A0219_10_27.bmp, -1, -1, -1
+A0219_10_28.bmp, -1, -1, -1
+A0219_10_29.bmp, -1, -1, -1
+A0219_10_30.bmp, -1, -1, -1
+A0219_10_31.bmp, -1, -1, -1
+A0219_10_32.bmp, -1, -1, -1
+A0219_10_34.bmp, -1, -1, -1
+A0219_10_35.bmp, -1, -1, -1
+A0219_10_36.bmp, -1, -1, -1
+A0219_10_37.bmp, -1, -1, -1
+A0219_10_38.bmp, -1, -1, -1
+A0219_10_39.bmp, -1, -1, -1
+A0219_10_40.bmp, -1, -1, -1
+A0219_10_41.bmp, -1, -1, -1
+A0219_10_42.bmp, -1, -1, -1
+A0219_10_43.bmp, -1, -1, -1
+A0219_10_44.bmp, -1, -1, -1
+A0219_10_46.bmp, -1, -1, -1
+A0219_10_47.bmp, -1, -1, -1
+A0219_10_48.bmp, -1, -1, -1
+A0219_10_49.bmp, -1, -1, -1
+A0219_10_50.bmp, -1, -1, -1
+A0219_10_51.bmp, -1, -1, -1
+A0219_10_52.bmp, -1, -1, -1
+A0219_10_53.bmp, -1, -1, -1
+A0219_10_54.bmp, -1, -1, -1
+A0219_10_55.bmp, -1, -1, -1
+A0219_10_56.bmp, -1, -1, -1
+A0219_10_58.bmp, -1, -1, -1
+A0219_10_61.bmp, -1, -1, -1
+A0219_10_62.bmp, -1, -1, -1
+A0219_10_63.bmp, -1, -1, -1
+A0219_10_64.bmp, -1, -1, -1
+A0219_10_65.bmp, -1, -1, -1
+A0219_10_66.bmp, -1, -1, -1
+A0219_10_67.bmp, -1, -1, -1
+A0219_10_68.bmp, -1, -1, -1
+A0219_10_69.bmp, -1, -1, -1
+A0219_10_70.bmp, -1, -1, -1
+A0219_10_71.bmp, -1, -1, -1
+A0219_10_72.bmp, -1, -1, -1
+A0219_10_73.bmp, -1, -1, -1
+A0219_10_74.bmp, -1, -1, -1
+A0219_10_75.bmp, -1, -1, -1
+A0235_10_00.bmp, -1, -1, -1
+A0235_10_01.bmp, -1, -1, -1
+A0235_10_02.bmp, -1, -1, -1
+A0235_10_03.bmp, -1, -1, -1
+A0235_10_04.bmp, -1, -1, -1
+A0235_10_05.bmp, -1, -1, -1
+A0235_10_06.bmp, -1, -1, -1
+A0235_10_07.bmp, -1, -1, -1
+A0235_10_08.bmp, -1, -1, -1
+A0235_10_10.bmp, -1, -1, -1
+A0235_10_11.bmp, -1, -1, -1
+A0235_10_12.bmp, -1, -1, -1
+A0235_10_14.bmp, -1, -1, -1
+A0235_10_15.bmp, -1, -1, -1
+A0235_10_16.bmp, -1, -1, -1
+A0235_10_17.bmp, -1, -1, -1
+A0235_10_18.bmp, -1, -1, -1
+A0235_10_19.bmp, -1, -1, -1
+A0235_10_20.bmp, -1, -1, -1
+A0235_10_21.bmp, -1, -1, -1
+A0235_10_24.bmp, -1, -1, -1
+A0235_10_26.bmp, -1, -1, -1
+A0235_10_27.bmp, -1, -1, -1
+A0235_10_28.bmp, -1, -1, -1
+A0235_10_29.bmp, -1, -1, -1
+A0235_10_30.bmp, -1, -1, -1
+A0235_10_31.bmp, -1, -1, -1
+A0235_10_32.bmp, -1, -1, -1
+A0235_10_34.bmp, -1, -1, -1
+A0235_10_35.bmp, -1, -1, -1
+A0235_10_36.bmp, -1, -1, -1
+A0235_10_37.bmp, -1, -1, -1
+A0235_10_38.bmp, -1, -1, -1
+A0235_10_39.bmp, -1, -1, -1
+A0235_10_40.bmp, -1, -1, -1
+A0235_10_41.bmp, -1, -1, -1
+A0235_10_42.bmp, -1, -1, -1
+A0235_10_43.bmp, -1, -1, -1
+A0235_10_44.bmp, -1, -1, -1
+A0235_10_46.bmp, -1, -1, -1
+A0235_10_47.bmp, -1, -1, -1
+A0235_10_48.bmp, -1, -1, -1
+A0235_10_49.bmp, -1, -1, -1
+A0235_10_50.bmp, -1, -1, -1
+A0235_10_51.bmp, -1, -1, -1
+A0235_10_52.bmp, -1, -1, -1
+A0235_10_53.bmp, -1, -1, -1
+A0235_10_54.bmp, -1, -1, -1
+A0235_10_55.bmp, -1, -1, -1
+A0235_10_56.bmp, -1, -1, -1
+A0235_10_58.bmp, -1, -1, -1
+A0235_10_61.bmp, -1, -1, -1
+A0235_10_62.bmp, -1, -1, -1
+A0235_10_63.bmp, -1, -1, -1
+A0235_10_64.bmp, -1, -1, -1
+A0235_10_65.bmp, -1, -1, -1
+A0235_10_66.bmp, -1, -1, -1
+A0235_10_67.bmp, -1, -1, -1
+A0235_10_68.bmp, -1, -1, -1
+A0235_10_69.bmp, -1, -1, -1
+A0235_10_70.bmp, -1, -1, -1
+A0235_10_71.bmp, -1, -1, -1
+A0235_10_72.bmp, -1, -1, -1
+A0235_10_73.bmp, -1, -1, -1
+A0235_10_74.bmp, -1, -1, -1
+A0235_10_75.bmp, -1, -1, -1
+A0249_10_00.bmp, -1, -1, -1
+A0249_10_01.bmp, -1, -1, -1
+A0249_10_02.bmp, -1, -1, -1
+A0249_10_03.bmp, -1, -1, -1
+A0249_10_04.bmp, -1, -1, -1
+A0249_10_05.bmp, -1, -1, -1
+A0249_10_06.bmp, -1, -1, -1
+A0249_10_07.bmp, -1, -1, -1
+A0249_10_08.bmp, -1, -1, -1
+A0249_10_10.bmp, -1, -1, -1
+A0249_10_11.bmp, -1, -1, -1
+A0249_10_12.bmp, -1, -1, -1
+A0249_10_14.bmp, -1, -1, -1
+A0249_10_15.bmp, -1, -1, -1
+A0249_10_16.bmp, -1, -1, -1
+A0249_10_17.bmp, -1, -1, -1
+A0249_10_18.bmp, -1, -1, -1
+A0249_10_19.bmp, -1, -1, -1
+A0249_10_20.bmp, -1, -1, -1
+A0249_10_21.bmp, -1, -1, -1
+A0249_10_24.bmp, -1, -1, -1
+A0249_10_26.bmp, -1, -1, -1
+A0249_10_27.bmp, -1, -1, -1
+A0249_10_28.bmp, -1, -1, -1
+A0249_10_29.bmp, -1, -1, -1
+A0249_10_30.bmp, -1, -1, -1
+A0249_10_31.bmp, -1, -1, -1
+A0249_10_32.bmp, -1, -1, -1
+A0249_10_34.bmp, -1, -1, -1
+A0249_10_35.bmp, -1, -1, -1
+A0249_10_36.bmp, -1, -1, -1
+A0249_10_37.bmp, -1, -1, -1
+A0249_10_38.bmp, -1, -1, -1
+A0249_10_39.bmp, -1, -1, -1
+A0249_10_40.bmp, -1, -1, -1
+A0249_10_41.bmp, -1, -1, -1
+A0249_10_42.bmp, -1, -1, -1
+A0249_10_43.bmp, -1, -1, -1
+A0249_10_44.bmp, -1, -1, -1
+A0249_10_46.bmp, -1, -1, -1
+A0249_10_47.bmp, -1, -1, -1
+A0249_10_48.bmp, -1, -1, -1
+A0249_10_49.bmp, -1, -1, -1
+A0249_10_50.bmp, -1, -1, -1
+A0249_10_51.bmp, -1, -1, -1
+A0249_10_52.bmp, -1, -1, -1
+A0249_10_53.bmp, -1, -1, -1
+A0249_10_54.bmp, -1, -1, -1
+A0249_10_55.bmp, -1, -1, -1
+A0249_10_56.bmp, -1, -1, -1
+A0249_10_58.bmp, -1, -1, -1
+A0249_10_61.bmp, -1, -1, -1
+A0249_10_62.bmp, -1, -1, -1
+A0249_10_63.bmp, -1, -1, -1
+A0249_10_64.bmp, -1, -1, -1
+A0249_10_65.bmp, -1, -1, -1
+A0249_10_66.bmp, -1, -1, -1
+A0249_10_67.bmp, -1, -1, -1
+A0249_10_68.bmp, -1, -1, -1
+A0249_10_69.bmp, -1, -1, -1
+A0249_10_70.bmp, -1, -1, -1
+A0249_10_71.bmp, -1, -1, -1
+A0249_10_72.bmp, -1, -1, -1
+A0249_10_73.bmp, -1, -1, -1
+A0249_10_74.bmp, -1, -1, -1
+A0249_10_75.bmp, -1, -1, -1

examplar_data_labels/YouTubeUGC/labels.txt

@@ -0,0 +1,1147 @@
+Sports_2160P-4e9f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.405
+Animation_480P-7a31_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.487
+CoverSong_480P-59f2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.418
+Sports_360P-32d3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.012
+LiveMusic_480P-38f3_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.973
+Animation_480P-4b86_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.495
+Sports_2160P-2568_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.401
+NewsClip_480P-6615_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.510
+Gaming_480P-6a5a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.468
+Vlog_2160P-62b2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.390
+NewsClip_360P-12fc_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.902
+Gaming_360P-0b98_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.140
+Lecture_720P-4c87_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.743
+Gaming_480P-61ee_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.631
+HowTo_360P-041c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.214
+CoverSong_1080P-0a86_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.228
+Gaming_480P-0dbc_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.973
+Sports_360P-1d5c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.781
+Vlog_1080P-1609_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.528
+Gaming_1080P-51fc_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.377
+HDR_2160P-664d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.753
+Animation_480P-35ee_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.290
+HowTo_720P-2cc6_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.814
+Animation_480P-073c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.407
+HDR_1080P-33e3_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.450
+Gaming_1080P-0ce6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.968
+Sports_2160P-3d85_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.517
+TelevisionClip_360P-74dd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.580
+Sports_2160P-1ddc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.466
+Sports_720P-5e39_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.685
+Animation_1080P-3e01_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.045
+Gaming_720P-0fdb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.263
+VerticalVideo_720P-665d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.023
+Animation_360P-7856_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.751
+Animation_1080P-209f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.189
+Sports_360P-11b7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.401
+Vlog_480P-5dfe_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.800
+NewsClip_480P-5a3b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.237
+LiveMusic_480P-4f88_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.267
+NewsClip_1080P-7816_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.855
+HowTo_360P-2074_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.536
+Lecture_480P-4bc3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.930
+Lecture_360P-3590_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.703
+LyricVideo_720P-59ed_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.315
+Animation_1080P-05f8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.901
+Gaming_480P-14fc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.822
+NewsClip_720P-6016_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.522
+HDR_2160P-4581_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.979
+MusicVideo_720P-7bcf_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.536
+VerticalVideo_1080P-6e19_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.840
+Vlog_360P-22a0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.286
+HowTo_480P-15c1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.264
+MusicVideo_480P-3c8b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.908
+Sports_2160P-49f1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.287
+LiveMusic_480P-0638_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.909
+LiveMusic_480P-58fb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.502
+HDR_2160P-15e2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.191
+Gaming_1080P-698a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.743
+CoverSong_1080P-0188_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.383
+Gaming_2160P-3a25_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.502
+TelevisionClip_1080P-3b9b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.251
+HDR_1080P-3a4a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.824
+Gaming_1080P-12d4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.494
+Gaming_480P-75f7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.798
+Gaming_2160P-6cd8_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.310
+Animation_720P-412a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.249
+CoverSong_720P-0239_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.520
+NewsClip_480P-606e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.921
+Sports_2160P-300d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.670
+VerticalVideo_360P-42f1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.603
+MusicVideo_480P-12fb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.059
+Sports_2160P-69b9_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.564
+VerticalVideo_480P-550d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.832
+Animation_360P-4edc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.682
+TelevisionClip_1080P-3e42_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.989
+CoverSong_360P-11f9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.790
+TelevisionClip_720P-1b61_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.299
+NewsClip_360P-311a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.447
+CoverSong_720P-6b8c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.988
+Vlog_2160P-6629_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.270
+LyricVideo_720P-068d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.825
+HowTo_480P-4b6a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.763
+Animation_1080P-646f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.376
+Sports_360P-3960_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.787
+CoverSong_360P-59d1_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.828
+VerticalVideo_1080P-2f00_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.360
+Vlog_360P-4697_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.551
+Vlog_480P-206f_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.779
+Sports_2160P-1733_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.373
+MusicVideo_1080P-0860_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.963
+Gaming_360P-3794_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.475
+VerticalVideo_720P-1ada_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.532
+CoverSong_1080P-3499_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.193
+CoverSong_480P-0170_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.118
+VerticalVideo_480P-790a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.248
+Gaming_720P-1cbb_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.010
+Animation_480P-1da7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.183
+Gaming_1080P-72c8_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.366
+Lecture_480P-71c0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.249
+VerticalVideo_360P-0d29_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.809
+Gaming_1080P-6dc6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.602
+LiveMusic_360P-6640_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.110
+Vlog_720P-6d56_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.776
+HowTo_720P-4d9e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.688
+LiveMusic_360P-5b57_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.773
+NewsClip_720P-4e87_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.368
+NewsClip_480P-7232_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.991
+HDR_1080P-206d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.791
+LiveMusic_360P-48d5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.325
+Animation_1080P-58d3_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.243
+LiveMusic_480P-6542_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.249
+HDR_1080P-49d6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.589
+Gaming_480P-0a03_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.676
+Lecture_720P-2f8c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.839
+NewsClip_480P-4e77_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.450
+Gaming_360P-48b0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.427
+TelevisionClip_720P-31ce_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.324
+Animation_720P-2525_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.730
+Lecture_480P-181a_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.785
+CoverSong_1080P-1963_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.882
+HDR_1080P-7825_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.514
+NewsClip_1080P-5b53_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.450
+Gaming_2160P-31f6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.311
+VerticalVideo_480P-572a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.499
+LiveMusic_720P-6343_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.304
+CoverSong_360P-2b4d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.796
+Gaming_1080P-5ae7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.328
+LiveMusic_1080P-3549_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.906
+VerticalVideo_360P-3936_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.794
+HowTo_480P-7c11_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.267
+Lecture_480P-02f2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.261
+Gaming_2160P-2436_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.473
+Vlog_720P-561e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.438
+Animation_1080P-2fbe_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.346
+Vlog_720P-372d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.047
+HowTo_1080P-7cf2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.717
+Vlog_480P-59dc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.206
+HowTo_1080P-7399_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.194
+MusicVideo_1080P-24f5_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.283
+Vlog_480P-7754_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.944
+MusicVideo_480P-0448_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.453
+TelevisionClip_1080P-3d10_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.052
+LiveMusic_360P-2508_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.004
+Sports_720P-058f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.280
+Gaming_1080P-4865_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.355
+LyricVideo_720P-739a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.877
+Gaming_480P-7ccb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.581
+Vlog_2160P-19f9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.919
+HowTo_720P-21c6_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.838
+Vlog_480P-6395_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.755
+VerticalVideo_360P-6490_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.973
+Animation_720P-7e7d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.589
+LyricVideo_480P-4346_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.193
+MusicVideo_720P-62df_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.246
+Sports_2160P-5ef3_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.209
+TelevisionClip_1080P-525d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.273
+TelevisionClip_480P-4c6b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.850
+CoverSong_480P-60a6_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.369
+CoverSong_720P-6d6a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.520
+Lecture_720P-6016_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.043
+Sports_480P-1056_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.149
+Vlog_360P-433e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.198
+HDR_1080P-35f4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.160
+Gaming_720P-469a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.124
+MusicVideo_360P-5f8d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.437
+Sports_2160P-2eda_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.307
+LyricVideo_480P-1484_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.145
+NewsClip_360P-66ae_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.218
+LyricVideo_1080P-0625_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.300
+Lecture_360P-30eb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.350
+Gaming_2160P-387f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.513
+Vlog_1080P-7062_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.944
+Vlog_2160P-2953_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.467
+Lecture_720P-2f38_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.552
+Sports_360P-3e68_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.508
+NewsClip_480P-543f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.423
+LiveMusic_720P-0d9f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.319
+LyricVideo_720P-2d24_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.199
+CoverSong_1080P-1b08_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.812
+Gaming_1080P-6530_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.920
+LiveMusic_480P-3d03_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.729
+Sports_720P-38a1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.165
+Lecture_1080P-0201_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.826
+Vlog_2160P-7b5c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.610
+Sports_480P-5f38_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.771
+Animation_1080P-2fff_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.369
+Sports_1080P-76a2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.054
+HDR_1080P-1be2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.756
+VerticalVideo_1080P-3d96_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.452
+Sports_2160P-391a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.281
+VerticalVideo_720P-0f61_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.522
+Vlog_1080P-18da_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.442
+Vlog_2160P-030a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.448
+Gaming_1080P-35fa_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.035
+VerticalVideo_1080P-1105_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.828
+CoverSong_360P-5a24_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.789
+LyricVideo_480P-0a5b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.247
+MusicVideo_360P-7b94_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.440
+Sports_480P-3195_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.537
+Sports_2160P-3794_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.158
+Lecture_360P-6d29_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.990
+MusicVideo_1080P-2d1c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.616
+LiveMusic_360P-3be0_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.720
+MusicVideo_360P-5699_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.776
+CoverSong_1080P-5430_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.969
+Gaming_1080P-190d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.898
+MusicVideo_360P-24b0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.512
+Vlog_720P-135c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.872
+Sports_1080P-08e1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.224
+NewsClip_480P-2642_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.688
+NewsClip_1080P-22b3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.812
+Gaming_720P-324d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.145
+Lecture_1080P-1969_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.649
+Sports_360P-1803_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.593
+HDR_1080P-548b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.626
+Sports_360P-6c6f_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.556
+Sports_480P-3404_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.533
+NewsClip_720P-5d89_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.480
+NewsClip_360P-1eae_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.467
+Vlog_720P-5d08_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.783
+HowTo_480P-56a5_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.085
+Gaming_480P-7a08_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.792
+Animation_720P-13b7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.318
+VerticalVideo_1080P-1c86_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.848
+LiveMusic_720P-653e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.112
+TelevisionClip_1080P-0604_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.765
+Sports_2160P-7a11_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.373
+TelevisionClip_1080P-6282_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.199
+CoverSong_720P-4006_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.414
+CoverSong_1080P-0f3a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.358
+LiveMusic_1080P-6fe2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.281
+Animation_1080P-6a33_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.052
+Vlog_720P-32b2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.365
+NewsClip_360P-1b1c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.514
+VerticalVideo_360P-2fa3_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.331
+MusicVideo_1080P-453f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.372
+Gaming_1080P-2927_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.441
+Animation_720P-7b29_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.862
+Lecture_360P-27db_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.092
+NewsClip_720P-6106_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.996
+Animation_2160P-6f3b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.082
+Vlog_480P-2cad_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.244
+Sports_2160P-279f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.526
+TelevisionClip_480P-59f0_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.896
+CoverSong_480P-64d0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.917
+HowTo_480P-470b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.858
+Lecture_480P-41b7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.328
+VerticalVideo_480P-2aa1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.402
+TelevisionClip_480P-19d3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.442
+HDR_1080P-6eb7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.046
+Sports_720P-3eb4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.802
+HDR_2160P-6eeb_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.402
+Vlog_2160P-6f92_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.886
+Animation_1080P-0c4f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.940
+Animation_720P-79ee_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.197
+TelevisionClip_720P-5e93_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.292
+Lecture_720P-1033_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.541
+Gaming_480P-4560_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.904
+Lecture_720P-07e0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.082
+HowTo_720P-06eb_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.130
+CoverSong_1080P-5e69_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.290
+Lecture_480P-74ea_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.258
+Gaming_720P-5973_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.067
+TelevisionClip_480P-436c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.823
+Sports_480P-0827_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.914
+Gaming_1080P-13e3_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.434
+LiveMusic_720P-3320_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.450
+HDR_2160P-4dcd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.763
+LiveMusic_1080P-28fe_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.693
+Vlog_1080P-52fe_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.838
+HowTo_720P-37d0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.429
+LiveMusic_1080P-6bbe_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.329
+Animation_1080P-18f5_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.158
+CoverSong_720P-7539_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.348
+Lecture_720P-11e7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.569
+LiveMusic_480P-4c3a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.907
+Gaming_1080P-223e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.283
+NewsClip_360P-22ce_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.698
+Lecture_480P-2ed0_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.453
+Animation_720P-4268_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.846
+MusicVideo_480P-3aa2_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.767
+NewsClip_480P-4072_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.481
+Vlog_720P-0d79_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.805
+Sports_1080P-1bf7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.490
+Vlog_1080P-62fc_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.255
+CoverSong_720P-6626_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.316
+HowTo_360P-127f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.332
+Animation_480P-6ef6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.607
+Sports_1080P-0063_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.447
+Gaming_720P-7afd_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.486
+NewsClip_360P-4288_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.393
+HowTo_1080P-36a9_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.427
+Lecture_720P-2b0f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.852
+Gaming_480P-6f4b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.670
+Sports_720P-00a1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.061
+MusicVideo_720P-2d7d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.988
+Gaming_360P-215f_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.869
+Gaming_360P-73c7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.247
+Lecture_360P-2276_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.742
+CoverSong_360P-4add_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.855
+MusicVideo_480P-5830_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.045
+Gaming_720P-4cda_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.620
+Sports_2160P-0455_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.287
+TelevisionClip_1080P-68c6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.561
+LiveMusic_480P-2a5e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.089
+Lecture_720P-003a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.812
+MusicVideo_480P-1eee_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.061
+Gaming_720P-2dbe_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.327
+Vlog_1080P-3744_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.948
+Gaming_2160P-34a1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.414
+MusicVideo_720P-4895_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.173
+NewsClip_360P-0ff8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.518
+Gaming_1080P-6d53_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.821
+MusicVideo_360P-4de7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.968
+Animation_360P-4b4c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.003
+Lecture_720P-1f22_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.233
+HowTo_480P-04f1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.454
+Lecture_480P-6191_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.502
+TelevisionClip_360P-7b23_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.997
+Vlog_480P-0ddc_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.075
+Vlog_2160P-7f05_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.111
+TelevisionClip_480P-2ead_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.371
+NewsClip_720P-6cde_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.297
+Gaming_1080P-777b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.050
+Sports_720P-69a0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.878
+Gaming_1080P-3bf7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.245
+Vlog_1080P-35cd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.196
+NewsClip_720P-6a19_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.547
+Vlog_2160P-56e4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.289
+Gaming_2160P-7e80_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.108
+Sports_480P-11ec_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.612
+TelevisionClip_720P-19de_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.346
+NewsClip_1080P-632e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.517
+CoverSong_360P-69ef_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.145
+Lecture_720P-53a1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.586
+Lecture_480P-11df_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.514
+MusicVideo_1080P-6c88_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.318
+LyricVideo_480P-3b2c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.472
+TelevisionClip_720P-7c06_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.396
+Vlog_2160P-09c9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.876
+NewsClip_480P-36d0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.764
+LyricVideo_1080P-5461_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.031
+LiveMusic_1080P-14af_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.450
+HowTo_1080P-52bb_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.272
+MusicVideo_480P-6026_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.033
+TelevisionClip_1080P-39e3_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.607
+Gaming_480P-7e76_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.397
+Gaming_360P-3eb6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.184
+Lecture_480P-71d6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.275
+Sports_480P-41a5_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.970
+LyricVideo_1080P-1d75_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.299
+HDR_2160P-6ed1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.724
+Vlog_1080P-2394_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.641
+HowTo_480P-7579_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.079
+VerticalVideo_480P-1bb9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.825
+Gaming_1080P-29b1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.829
+NewsClip_360P-0376_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.095
+HowTo_720P-111e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.938
+LyricVideo_480P-3ccf_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.639
+Gaming_1080P-45fc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.695
+Gaming_2160P-4139_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.923
+Lecture_360P-6656_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.458
+LyricVideo_720P-4253_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.066
+Lecture_720P-3b7f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.106
+LyricVideo_720P-5d3a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.939
+Vlog_480P-535d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.846
+HowTo_360P-7fb1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.322
+HowTo_720P-12f7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.300
+Sports_720P-2c06_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.978
+HowTo_1080P-13aa_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.336
+Sports_720P-7ba7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.566
+Vlog_1080P-1e70_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.442
+Vlog_2160P-342a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.321
+Animation_360P-631c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.236
+NewsClip_480P-15fa_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.581
+HowTo_720P-0c47_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.364
+Sports_720P-3338_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.483
+Sports_360P-5ded_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.784
+NewsClip_1080P-67dc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.977
+Gaming_480P-6548_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.871
+LyricVideo_480P-200a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.063
+LyricVideo_720P-1c89_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.930
+NewsClip_1080P-3c7c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.462
+CoverSong_1080P-3409_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.331
+Gaming_1080P-6e45_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.517
+LiveMusic_720P-65f6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.019
+HowTo_720P-0b01_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.111
+Gaming_360P-2164_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.245
+Vlog_2160P-3019_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.578
+MusicVideo_360P-17e4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.233
+Vlog_2160P-4419_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.324
+TelevisionClip_1080P-5e68_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.604
+HDR_1080P-2d32_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.503
+Lecture_480P-369f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.193
+Animation_480P-66c0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.453
+CoverSong_480P-3e4b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.877
+Sports_1080P-0d0c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.980
+Lecture_480P-7205_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.417
+LyricVideo_720P-6fc9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.671
+HowTo_480P-2610_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.981
+NewsClip_720P-1971_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.272
+Lecture_360P-7a57_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.493
+Sports_360P-02c3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.396
+Gaming_720P-0fba_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.089
+NewsClip_360P-1e1c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.254
+Sports_360P-6f62_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.594
+VerticalVideo_480P-3a6a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.337
+Sports_480P-5224_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.866
+Gaming_2160P-3ce4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.529
+CoverSong_1080P-5cae_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.209
+Gaming_480P-6cdc_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.425
+Lecture_360P-7f7e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.679
+HowTo_1080P-0267_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.234
+Gaming_360P-043e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.582
+Sports_720P-17fb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.591
+VerticalVideo_720P-7859_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.603
+LyricVideo_480P-2c50_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.985
+Vlog_2160P-522f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.324
+Lecture_360P-03bc_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.762
+Gaming_480P-0d23_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.466
+Gaming_1080P-0ef8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.781
+Gaming_480P-09ab_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.795
+Sports_1080P-28a6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.452
+Vlog_1080P-1a53_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.474
+HowTo_480P-0d00_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.998
+Gaming_1080P-6578_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.399
+MusicVideo_1080P-55af_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.748
+Sports_720P-2191_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.789
+VerticalVideo_2160P-6d62_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.199
+MusicVideo_1080P-65c6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.066
+Vlog_720P-11c5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.816
+Animation_2160P-41dc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.512
+Sports_720P-5ea4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.238
+MusicVideo_720P-4ad2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.756
+CoverSong_1080P-5456_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.736
+CoverSong_480P-53f4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.169
+Sports_1080P-7584_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.755
+Vlog_360P-4795_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.479
+HowTo_360P-7dcd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.035
+Vlog_1080P-5904_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.365
+MusicVideo_480P-5461_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.272
+Gaming_720P-221d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.252
+Vlog_1080P-45c9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.888
+MusicVideo_360P-5f07_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.921
+VerticalVideo_480P-2fa4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.133
+Sports_1080P-43e2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.414
+HowTo_480P-32c9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.510
+Sports_720P-0b9e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.942
+Animation_720P-0116_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.054
+Gaming_720P-25aa_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.952
+Vlog_2160P-700c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.881
+Animation_360P-3e52_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.651
+LyricVideo_1080P-2946_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.177
+CoverSong_1080P-3df8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.827
+Gaming_2160P-348d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.443
+VerticalVideo_1080P-766b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.305
+Animation_480P-4e36_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.303
+VerticalVideo_480P-51b7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.299
+TelevisionClip_480P-3284_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.169
+HDR_2160P-6fab_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.443
+LyricVideo_1080P-0075_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.814
+Sports_720P-4c6f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.105
+Lecture_360P-2d1f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.620
+LiveMusic_1080P-2f7f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.326
+LyricVideo_480P-6fca_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.572
+Vlog_2160P-4655_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.887
+Lecture_720P-6951_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.163
+TelevisionClip_480P-373d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.044
+NewsClip_720P-72d2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.357
+LiveMusic_360P-3b96_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.358
+Sports_360P-2ace_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.444
+CoverSong_480P-2142_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.250
+Lecture_480P-42c3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.451
+Lecture_720P-0f64_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.580
+HowTo_480P-4948_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.551
+CoverSong_720P-449f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.496
+LyricVideo_720P-0940_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.045
+Vlog_1080P-34cb_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.428
+Vlog_1080P-7e8c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.042
+LyricVideo_360P-5e87_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.999
+Gaming_480P-5a5a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.478
+CoverSong_480P-1109_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.826
+Animation_480P-3fdf_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.465
+MusicVideo_720P-14f0_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.382
+Gaming_1080P-173a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.013
+Animation_720P-06a6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.942
+Gaming_360P-2330_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.607
+Sports_720P-531c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.555
+NewsClip_480P-437c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.011
+Sports_360P-27b0_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.820
+VerticalVideo_360P-7ba8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.563
+Animation_720P-2abf_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.615
+Lecture_480P-4272_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.457
+Lecture_480P-3bdc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.168
+HDR_1080P-0d31_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.373
+CoverSong_720P-10f1_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.805
+TelevisionClip_1080P-5278_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.099
+CoverSong_1080P-237f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.893
+TelevisionClip_480P-27ca_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.475
+Gaming_1080P-2e97_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.391
+NewsClip_720P-739b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.007
+VerticalVideo_720P-7c1d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.502
+HowTo_720P-7878_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.674
+Vlog_480P-6c60_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.487
+Animation_720P-3adc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.575
+Lecture_480P-1ec3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.186
+Sports_480P-7f7e_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.953
+TelevisionClip_480P-3617_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.312
+MusicVideo_480P-66b2_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.709
+NewsClip_480P-31bd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.610
+LiveMusic_360P-2b0f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.401
+LiveMusic_480P-2019_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.631
+Animation_480P-046c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.150
+Vlog_720P-329f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.700
+Gaming_1080P-564e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.322
+Animation_360P-69e0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.382
+MusicVideo_720P-0752_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.365
+LiveMusic_1080P-7948_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.838
+CoverSong_360P-6d6a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.273
+HowTo_1080P-6cbf_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.915
+Vlog_1080P-4f26_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.962
+Sports_360P-0c66_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.097
+Lecture_360P-506c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.168
+MusicVideo_1080P-04b6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.288
+HDR_1080P-3521_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.895
+Sports_360P-6b4f_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.981
+Animation_720P-620f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.522
+Sports_2160P-6b28_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.391
+MusicVideo_480P-6fb6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.611
+Lecture_720P-10bc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.116
+Sports_1080P-3a3b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.669
+NewsClip_1080P-4241_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.936
+LiveMusic_360P-7483_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.063
+HDR_2160P-70ca_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.905
+Animation_720P-41d6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.148
+CoverSong_360P-0aa2_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.511
+TelevisionClip_480P-3c64_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.456
+Sports_2160P-2a83_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.002
+Vlog_2160P-255c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.443
+CoverSong_480P-5b62_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.056
+Gaming_720P-5ba2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.281
+VerticalVideo_360P-7ec3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.909
+Animation_1080P-4be3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.565
+LiveMusic_480P-61ef_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.153
+NewsClip_360P-7a96_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.291
+Sports_480P-4cf8_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.023
+Gaming_480P-0109_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.436
+Animation_360P-794f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.362
+Vlog_2160P-7bfb_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.206
+CoverSong_480P-2d2c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.837
+VerticalVideo_720P-6580_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.979
+NewsClip_480P-696e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.102
+VerticalVideo_720P-7517_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.271
+Gaming_720P-312f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.826
+TelevisionClip_480P-723e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.882
+CoverSong_480P-3e92_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.235
+LyricVideo_1080P-16b6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.381
+VerticalVideo_360P-1424_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.732
+HowTo_720P-0a67_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.315
+Gaming_2160P-2cb1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.481
+NewsClip_720P-672c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.987
+NewsClip_720P-04ba_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.490
+CoverSong_360P-5d20_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.943
+Vlog_2160P-3a75_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.212
+Vlog_360P-6279_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.260
+Vlog_360P-1e5b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.266
+VerticalVideo_1080P-3a9b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.432
+Sports_480P-44e5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.170
+HDR_2160P-06ae_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.474
+Sports_480P-35ea_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.668
+MusicVideo_720P-3c4c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.135
+Vlog_1080P-19bb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.076
+NewsClip_360P-439a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.282
+Vlog_720P-03d5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.605
+Gaming_2160P-5c53_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.309
+TelevisionClip_1080P-401e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.227
+CoverSong_480P-35e1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.608
+Sports_2160P-7af7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.950
+Sports_360P-5252_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.538
+Sports_480P-3ca4_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.389
+Vlog_2160P-13a1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.859
+Gaming_2160P-4258_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.067
+Animation_1080P-0cdf_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.245
+Lecture_360P-4f00_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.172
+Animation_720P-0acc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.876
+CoverSong_360P-67ec_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.404
+VerticalVideo_360P-634f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.523
+NewsClip_360P-5f3b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.298
+MusicVideo_360P-462f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.510
+Sports_360P-50fd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.011
+CoverSong_1080P-1b0c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.763
+Sports_720P-3ffe_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.011
+Sports_1080P-4e05_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.445
+Vlog_2160P-1ff6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.290
+Gaming_1080P-6db2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.475
+Gaming_480P-3a44_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.043
+NewsClip_480P-0ce5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.265
+Sports_360P-4802_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.582
+Sports_720P-2632_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.884
+NewsClip_480P-2b9b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.421
+HDR_1080P-687e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.629
+Animation_480P-791b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.286
+HowTo_360P-09f8_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.951
+MusicVideo_720P-575d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.242
+Vlog_480P-1b39_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.325
+HDR_1080P-601b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.318
+Sports_2160P-210c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.462
+HowTo_1080P-63ec_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.229
+HowTo_360P-4276_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.448
+Vlog_480P-4beb_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.512
+NewsClip_720P-579b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.866
+LyricVideo_1080P-1b04_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.803
+HowTo_720P-6791_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.263
+LiveMusic_360P-54d0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.559
+Vlog_480P-279d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.373
+VerticalVideo_1080P-2195_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.891
+NewsClip_1080P-00c3_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.421
+HowTo_1080P-03fd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.517
+NewsClip_360P-2986_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.902
+NewsClip_1080P-06df_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.489
+LyricVideo_1080P-4b40_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.636
+CoverSong_1080P-2207_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.675
+MusicVideo_1080P-16e6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.349
+Sports_2160P-7165_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.351
+CoverSong_720P-6d81_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.145
+MusicVideo_720P-5c9c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.471
+Gaming_480P-1542_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.776
+Vlog_360P-7efe_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.912
+LyricVideo_720P-09ed_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.026
+HowTo_360P-6a0e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.104
+HDR_2160P-3663_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.340
+LiveMusic_720P-0cd5_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.042
+VerticalVideo_720P-4ca7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.442
+VerticalVideo_360P-145e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.606
+Lecture_360P-5aea_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.088
+LiveMusic_360P-6a65_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.427
+HowTo_360P-2a3b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.589
+Sports_480P-2dfe_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.401
+LiveMusic_1080P-21dd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.819
+Sports_1080P-7dba_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.718
+Vlog_720P-343d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.667
+Sports_480P-0af9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.765
+Sports_2160P-4201_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.012
+Sports_480P-0623_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.106
+HowTo_720P-7782_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.761
+VerticalVideo_720P-4730_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.107
+NewsClip_360P-5752_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.346
+Sports_2160P-086d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.376
+LyricVideo_480P-0f72_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.064
+Vlog_1080P-64b6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.192
+LiveMusic_1080P-3f95_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.050
+TelevisionClip_480P-09d8_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.908
+Sports_1080P-19d8_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.472
+Gaming_720P-6625_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.813
+MusicVideo_1080P-7706_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.925
+LyricVideo_360P-3afc_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.685
+NewsClip_720P-7e56_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.410
+HowTo_1080P-1f3b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.185
+NewsClip_480P-2ba7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.791
+Vlog_480P-7d0c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.577
+Lecture_480P-6b1e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.368
+Sports_480P-6508_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.448
+VerticalVideo_720P-0750_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.710
+NewsClip_720P-7b97_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.803
+LyricVideo_720P-47a9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.816
+Lecture_720P-3f9f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.479
+Sports_2160P-2e1d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.457
+LiveMusic_360P-6266_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.383
+VerticalVideo_360P-2bb0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.073
+Animation_720P-01b3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.638
+MusicVideo_480P-184c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.726
+TelevisionClip_720P-3f4c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.515
+Sports_2160P-1b70_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.508
+Gaming_360P-63e6_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.636
+Sports_2160P-4024_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.054
+Sports_360P-2725_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.862
+Vlog_1080P-1df9_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.134
+Gaming_2160P-2b92_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.632
+CoverSong_360P-14b9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.633
+Gaming_360P-586d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.026
+Gaming_720P-40b2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.855
+Gaming_720P-6658_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.715
+TelevisionClip_720P-44d1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.839
+Gaming_720P-64b0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.396
+MusicVideo_1080P-0706_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.084
+Vlog_1080P-5f28_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.811
+Sports_720P-5ae1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.333
+VerticalVideo_1080P-7cf3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.955
+Sports_1080P-2a21_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.231
+Animation_360P-08c9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.652
+NewsClip_720P-7c67_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.434
+NewsClip_360P-67ce_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.196
+Animation_1080P-21dd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.791
+LyricVideo_1080P-63d9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.512
+MusicVideo_1080P-5072_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.190
+Sports_360P-65f3_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.826
+NewsClip_720P-2882_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.667
+CoverSong_480P-6c3e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.502
+Lecture_480P-7d77_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.598
+Lecture_360P-7550_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.033
+Gaming_2160P-6bb6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.537
+Vlog_720P-5364_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.794
+Lecture_480P-2513_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.879
+HowTo_720P-4d55_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.947
+MusicVideo_480P-2de0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.470
+Animation_480P-087e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.525
+NewsClip_1080P-3427_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.382
+Vlog_1080P-010b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.185
+NewsClip_720P-0c81_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.893
+Sports_360P-4ad7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.327
+TelevisionClip_480P-1b49_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.183
+LyricVideo_720P-36fb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.839
+NewsClip_480P-4a9f_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.983
+NewsClip_480P-6bfa_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.793
+MusicVideo_1080P-18f5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.829
+Sports_1080P-6710_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.335
+NewsClip_360P-5bcc_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.864
+Gaming_1080P-3d58_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.782
+HowTo_720P-017a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.762
+Gaming_480P-6491_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.013
+HowTo_360P-18e7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.780
+LyricVideo_480P-7919_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.508
+NewsClip_480P-7a0d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.754
+HowTo_480P-60b3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.576
+MusicVideo_360P-648c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.272
+LiveMusic_1080P-2930_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.776
+TelevisionClip_360P-11d5_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.842
+Lecture_720P-094d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.616
+Gaming_1080P-2221_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.239
+CoverSong_480P-3076_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.710
+HDR_1080P-3181_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.777
+CoverSong_1080P-5f92_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.043
+TelevisionClip_360P-0951_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.511
+Sports_2160P-4aae_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.323
+Lecture_720P-5c7e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.846
+Gaming_1080P-7a1e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.024
+HDR_1080P-13eb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.761
+LiveMusic_720P-2620_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.721
+MusicVideo_1080P-3a50_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.690
+Vlog_1080P-6686_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.926
+Sports_480P-211f_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.715
+Animation_1080P-5083_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.252
+Lecture_1080P-0c8a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.550
+TelevisionClip_1080P-63e6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.228
+Gaming_720P-493e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.567
+NewsClip_360P-0c17_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.968
+Gaming_480P-626a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.786
+Lecture_720P-1e62_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.958
+CoverSong_720P-60d3_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.991
+Sports_2160P-349c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.302
+Animation_480P-0d04_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.111
+Vlog_720P-033a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.302
+HowTo_720P-269e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.643
+VerticalVideo_1080P-4925_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.496
+VerticalVideo_1080P-7c5c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.780
+VerticalVideo_1080P-4b92_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.119
+Vlog_2160P-77d8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.667
+Gaming_720P-6a45_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.967
+LiveMusic_360P-5281_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.596
+LiveMusic_720P-58db_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.382
+Gaming_1080P-71a5_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.482
+MusicVideo_720P-44c1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.499
+Lecture_720P-5725_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.262
+Lecture_360P-114f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.172
+MusicVideo_1080P-106d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.259
+LiveMusic_1080P-77e8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.346
+Vlog_480P-5e1c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.581
+HowTo_720P-7c38_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.074
+NewsClip_480P-0269_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.040
+TelevisionClip_1080P-4c24_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.334
+MusicVideo_1080P-7265_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.566
+LiveMusic_360P-3979_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.838
+Vlog_2160P-0577_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.270
+LyricVideo_1080P-41ee_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.071
+NewsClip_720P-4603_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.454
+HDR_1080P-4f4a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.510
+HowTo_360P-06be_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.601
+HDR_2160P-5926_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.386
+HowTo_480P-221e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.283
+Animation_720P-7ff2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.124
+Vlog_2160P-2b2d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.906
+VerticalVideo_720P-2efc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.523
+Sports_1080P-7203_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.910
+Gaming_2160P-416f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.332
+MusicVideo_720P-0355_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.907
+LyricVideo_720P-74a0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.415
+VerticalVideo_480P-34b9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.527
+Sports_720P-2234_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.160
+HowTo_480P-0eb3_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.507
+Vlog_480P-476b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.431
+Gaming_2160P-3aec_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.646
+Sports_2160P-7af8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.563
+VerticalVideo_720P-669d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.176
+NewsClip_1080P-0abf_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.672
+Animation_360P-5712_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.552
+LiveMusic_480P-459c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.755
+Gaming_360P-5e0f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.012
+MusicVideo_1080P-2b2b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.436
+Gaming_1080P-57ca_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.409
+Sports_480P-5871_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.698
+HowTo_480P-4a28_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.659
+Sports_2160P-3a9a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.633
+MusicVideo_1080P-34ff_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.658
+HowTo_360P-2fd5_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.049
+LyricVideo_360P-0ac7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.964
+MusicVideo_360P-5578_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.873
+MusicVideo_480P-41ce_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.766
+Lecture_360P-5d0a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.416
+Vlog_720P-155f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.316
+MusicVideo_1080P-7f2e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.750
+NewsClip_1080P-5be1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.065
+CoverSong_480P-7f6d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.048
+CoverSong_360P-13d2_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.578
+LiveMusic_480P-1a91_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.173
+CoverSong_1080P-0cdc_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.741
+CoverSong_360P-1b2b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.409
+Lecture_480P-4c86_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.666
+Sports_720P-50cf_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.372
+Animation_480P-52af_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.656
+HowTo_480P-63a2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.559
+NewsClip_480P-2407_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.688
+Vlog_2160P-7324_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.194
+Sports_720P-4b69_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.903
+HDR_1080P-3749_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.080
+HDR_2160P-40ab_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.105
+Gaming_480P-6d1e_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.674
+Sports_1080P-6571_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.652
+LiveMusic_1080P-541f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.174
+Gaming_720P-3524_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.110
+HowTo_1080P-7f18_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.718
+Animation_1080P-4ea8_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.121
+LyricVideo_720P-6f0c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.803
+Lecture_360P-20c3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.133
+TelevisionClip_1080P-6026_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.415
+Lecture_480P-5aee_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.606
+Lecture_480P-74cf_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.428
+Vlog_480P-34ba_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.666
+Vlog_2160P-327a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.983
+Gaming_1080P-58ba_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.952
+LiveMusic_720P-267b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.525
+Lecture_480P-2655_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.261
+Lecture_480P-5cd7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.024
+Vlog_2160P-70d6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.067
+Gaming_720P-103a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.056
+Lecture_1080P-1709_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.172
+HowTo_1080P-64f7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.454
+HowTo_720P-3a5d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.536
+NewsClip_720P-35d9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.017
+Gaming_360P-21d2_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.372
+CoverSong_1080P-0f33_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.221
+Vlog_480P-7237_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.890
+Sports_480P-19e4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.835
+VerticalVideo_360P-6ff2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.543
+Vlog_1080P-687a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.374
+Sports_1080P-53a0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.429
+HowTo_1080P-63e4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.948
+NewsClip_1080P-2eb0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.484
+Sports_1080P-1d78_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.378
+HowTo_1080P-4d7b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.199
+Sports_480P-1396_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.459
+CoverSong_1080P-3aac_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.381
+Animation_480P-6e23_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.731
+NewsClip_720P-5564_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.377
+Sports_1080P-49c5_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.142
+MusicVideo_480P-7643_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.093
+HowTo_480P-3435_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.930
+MusicVideo_720P-3b11_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.600
+NewsClip_1080P-1db0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.141
+LiveMusic_1080P-6d1a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.780
+HDR_2160P-2a72_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.113
+Vlog_2160P-4f98_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.470
+VerticalVideo_720P-19fa_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.269
+Sports_720P-6bb7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.882
+NewsClip_720P-7b7e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.145
+NewsClip_360P-3451_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.121
+HowTo_360P-3aa6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.019
+Sports_1080P-241e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.411
+Vlog_360P-76ae_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.606
+VerticalVideo_480P-7278_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.005
+Gaming_2160P-67b0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.204
+CoverSong_360P-3218_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.427
+LyricVideo_1080P-756a_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.228
+LiveMusic_480P-474a_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.491
+VerticalVideo_360P-54f7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.146
+MusicVideo_1080P-4671_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.083
+Lecture_360P-5779_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.069
+VerticalVideo_1080P-1ac1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.979
+Gaming_1080P-277c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.884
+Sports_720P-62ba_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.965
+Sports_360P-4545_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.432
+Lecture_360P-2c55_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.717
+LyricVideo_360P-17ce_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.830
+HowTo_1080P-763c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.957
+Sports_360P-2e48_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.744
+Sports_1080P-3eb0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.819
+Gaming_720P-6403_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.023
+HowTo_360P-0562_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.261
+Gaming_360P-56fe_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.439
+HDR_1080P-69de_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.181
+Lecture_1080P-011f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.590
+Vlog_360P-3bc2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.944
+Animation_720P-41cf_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.862
+LiveMusic_720P-289b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.436
+HowTo_360P-4e4c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.005
+Lecture_720P-7e40_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.246
+Gaming_2160P-5c33_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.834
+Gaming_1080P-66d2_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.142
+TelevisionClip_1080P-3758_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.313
+Lecture_720P-2442_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.390
+NewsClip_1080P-02fd_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.159
+HowTo_480P-4c99_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.940
+NewsClip_480P-28eb_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.998
+Vlog_720P-3e9c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.882
+Sports_1080P-679d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.366
+NewsClip_720P-5787_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.483
+LiveMusic_360P-22c5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.231
+HDR_2160P-5e25_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.209
+Lecture_480P-7e55_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.538
+Vlog_480P-08c7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.591
+MusicVideo_1080P-6fb1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.492
+HowTo_720P-483c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.485
+NewsClip_480P-72b8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.407
+CoverSong_720P-1393_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.059
+Vlog_2160P-5874_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.101
+MusicVideo_360P-2fcb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.619
+Vlog_2160P-13fe_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.992
+Sports_720P-2c80_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.136
+Sports_1080P-1ae3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.352
+NewsClip_480P-49c9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.975
+VerticalVideo_360P-579c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.058
+NewsClip_360P-274e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.332
+Sports_480P-3dfd_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.498
+NewsClip_1080P-27fb_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.002
+MusicVideo_720P-3698_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.378
+MusicVideo_720P-734f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.848
+Vlog_1080P-1f0a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.872
+Vlog_2160P-416c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.019
+Animation_360P-47cc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.282
+HowTo_360P-6093_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.205
+NewsClip_480P-41b1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.623
+Vlog_1080P-21f5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.327
+Sports_720P-5bfd_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.677
+CoverSong_720P-5ba2_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.834
+Gaming_1080P-26dc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.541
+CoverSong_720P-5689_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.317
+LiveMusic_720P-66df_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.592
+Animation_480P-70d4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.848
+MusicVideo_360P-5358_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.069
+LyricVideo_480P-51e4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.179
+LiveMusic_720P-4ae2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.285
+VerticalVideo_1080P-3709_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.367
+Sports_1080P-47e9_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.106
+Vlog_1080P-23cb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.069
+Lecture_720P-5120_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.295
+Gaming_360P-7975_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.761
+NewsClip_480P-3672_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.802
+MusicVideo_360P-08cb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.401
+Animation_720P-6372_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.422
+Sports_720P-675b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.201
+LiveMusic_1080P-59b3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.443
+TelevisionClip_360P-3b9a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.883
+Lecture_720P-50b9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.579
+Gaming_360P-6207_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.349
+Gaming_2160P-2dc4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.383
+HowTo_360P-55e9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.496
+NewsClip_360P-4417_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.924
+HowTo_480P-118d_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.462
+Gaming_2160P-28de_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.349
+Sports_480P-3f50_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.234
+Sports_480P-0efe_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.673
+MusicVideo_480P-4cc8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.346
+Vlog_720P-141f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.128
+Sports_480P-1fe9_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.706
+Gaming_1080P-1704_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.295
+Animation_360P-5de0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.429
+TelevisionClip_360P-29f1_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.581
+VerticalVideo_1080P-04d4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.150
+Sports_2160P-1261_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.114
+Sports_1080P-5d25_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.478
+LyricVideo_480P-5c17_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.344
+Animation_360P-24d4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.435
+NewsClip_360P-5052_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.095
+Sports_720P-6d04_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.783
+TelevisionClip_480P-30b6_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.609
+Animation_720P-1a6d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.283
+Lecture_360P-311d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.603
+Gaming_360P-4f09_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.506
+Vlog_1080P-26dc_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.685
+Gaming_360P-2f99_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.281
+Vlog_1080P-2cda_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.429
+VerticalVideo_720P-456c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.711
+Sports_2160P-7bbe_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.343
+VerticalVideo_480P-419c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.780
+Sports_1080P-2524_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.662
+HDR_2160P-5275_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.206
+VerticalVideo_360P-52c1_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.352
+LiveMusic_480P-65ca_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.691
+NewsClip_1080P-48ae_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.887
+LiveMusic_360P-265c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.339
+LiveMusic_360P-121f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.430
+Gaming_1080P-3a9d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.872
+Lecture_360P-4bb4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.822
+MusicVideo_480P-7955_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.509
+Animation_1080P-3dbf_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.444
+VerticalVideo_1080P-73b6_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.034
+HDR_1080P-1e5b_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.242
+Lecture_720P-7189_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.869
+MusicVideo_480P-61ba_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.756
+Sports_720P-33c6_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.985
+LyricVideo_1080P-584f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.188
+LyricVideo_720P-0ae4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.717
+CoverSong_480P-41a5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.270
+Sports_360P-61f6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.393
+Sports_480P-1019_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.316
+Animation_480P-08a5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.420
+Vlog_2160P-310b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.857
+Gaming_480P-20ca_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.891
+Animation_720P-57d9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.522
+Lecture_480P-7eec_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.015
+MusicVideo_1080P-6260_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.042
+Gaming_480P-445b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.714
+Gaming_2160P-7ff0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.238
+TelevisionClip_720P-4edb_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.015
+HowTo_480P-0cb3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.592
+Animation_720P-431c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.396
+CoverSong_720P-7360_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.798
+Vlog_2160P-408f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.790
+MusicVideo_1080P-1a53_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.577
+VerticalVideo_360P-3b4e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.576
+Gaming_360P-187a_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.133
+Sports_1080P-3db7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.021
+MusicVideo_480P-483b_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.539
+VerticalVideo_1080P-4da5_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.257
+VerticalVideo_360P-5429_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.616
+Sports_720P-07d0_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.863
+HDR_1080P-55c4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.106
+Gaming_360P-7acb_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.059
+LiveMusic_480P-559d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.695
+Gaming_2160P-3002_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.337
+Vlog_720P-2929_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.918
+Vlog_480P-0980_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.755
+Sports_480P-77e3_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.053
+Gaming_1080P-44c4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.839
+CoverSong_1080P-033d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.433
+HDR_1080P-549e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.064
+VerticalVideo_1080P-34ba_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.382
+Vlog_720P-60f8_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.335
+Sports_1080P-46ed_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.493
+VerticalVideo_720P-42f2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.199
+Gaming_360P-279f_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.841
+CoverSong_480P-1019_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.815
+Vlog_360P-3121_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.775
+Sports_480P-0bb9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.419
+Sports_1080P-15d1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.570
+TelevisionClip_480P-0e46_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.881
+CoverSong_360P-6da7_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.763
+Lecture_360P-5594_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.367
+HowTo_360P-2d35_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.847
+Sports_480P-2053_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.149
+Lecture_360P-051a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.640
+LyricVideo_720P-7f44_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.196
+Lecture_480P-5f3a_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.324
+CoverSong_1080P-6f26_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.288
+NewsClip_360P-2c91_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.859
+CoverSong_360P-2146_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.273
+CoverSong_360P-53a6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.235
+LiveMusic_1080P-7ead_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.559
+Vlog_2160P-059c_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.441
+Sports_360P-301d_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.667
+Sports_720P-3072_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.180
+HowTo_720P-3813_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.072
+Sports_720P-5833_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.205
+TelevisionClip_720P-1862_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.238
+HowTo_1080P-55d1_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.102
+NewsClip_720P-7745_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.528
+VerticalVideo_720P-0dac_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.934
+Sports_2160P-324e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.317
+Sports_2160P-2626_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.388
+LiveMusic_1080P-1ace_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.162
+LyricVideo_360P-5868_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.107
+CoverSong_1080P-4bad_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.298
+Vlog_2160P-4362_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.554
+NewsClip_360P-1093_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.100
+VerticalVideo_1080P-4591_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.439
+LyricVideo_1080P-63d4_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.221
+Animation_360P-3e40_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.353
+HowTo_360P-5da7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.425
+HDR_1080P-68c8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.804
+Gaming_480P-6c92_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.348
+Vlog_2160P-7b10_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.864
+Vlog_480P-5275_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.120
+Animation_1080P-5dd8_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.958
+NewsClip_480P-5e7f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.105
+VerticalVideo_360P-70a9_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.814
+CoverSong_480P-2207_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.934
+Vlog_720P-03f9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.972
+CoverSong_480P-3f3b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.027
+Lecture_360P-44c2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.389
+CoverSong_720P-2d05_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.224
+MusicVideo_1080P-6c1d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.323
+HowTo_480P-64a2_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.836
+TelevisionClip_720P-4af1_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.213
+HowTo_720P-6323_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.732
+Vlog_1080P-4ba9_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.908
+CoverSong_480P-4d34_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.677
+Vlog_720P-1003_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.604
+HowTo_720P-479b_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.620
+TelevisionClip_1080P-3d83_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.382
+Sports_1080P-0640_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.524
+Gaming_1080P-45af_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.373
+Animation_480P-7ef2_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.164
+Vlog_480P-7615_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.588
+TelevisionClip_480P-415c_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.544
+Vlog_720P-4e3d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.743
+Sports_360P-0dda_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.868
+LyricVideo_1080P-3803_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.950
+Sports_480P-1d1b_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.177
+Vlog_1080P-25de_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.914
+CoverSong_480P-6c50_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.729
+LiveMusic_1080P-51f6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.819
+Sports_480P-6e41_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.306
+Sports_720P-0104_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.223
+LiveMusic_360P-1d94_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.028
+Sports_360P-7f50_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.963
+Vlog_360P-2973_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.902
+HowTo_360P-1dba_crf_10_ss_00_t_20.0.mp4, -1, -1, 1.907
+VerticalVideo_1080P-360f_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.189
+NewsClip_720P-2182_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.494
+Lecture_480P-73f6_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.256
+Animation_1080P-3d67_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.778
+VerticalVideo_480P-467e_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.383
+NewsClip_720P-23e0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.110
+Animation_360P-188f_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.702
+CoverSong_720P-014c_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.618
+NewsClip_720P-37f7_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.716
+Animation_1080P-4214_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.510
+VerticalVideo_720P-6bf7_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.261
+LiveMusic_720P-6452_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.098
+HDR_1080P-46a4_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.582
+LiveMusic_1080P-157b_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.360
+Vlog_360P-2e9d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.746
+Animation_1080P-6ec0_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.124
+CoverSong_360P-0a9d_crf_10_ss_00_t_20.0.mp4, -1, -1, 2.538
+Gaming_720P-4813_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.290
+LyricVideo_1080P-725e_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.399
+LiveMusic_720P-71c5_crf_10_ss_00_t_20.0.mp4, -1, -1, 3.605
+Gaming_2160P-673d_crf_10_ss_00_t_20.0.mp4, -1, -1, 4.165

requirements.txt

@@ -0,0 +1,15 @@
+torch~=1.13
+torchvision
+opencv-python
+decord
+matplotlib
+scipy
+numpy
+tqdm
+timm
+einops
+wandb
+scikit-video
+thop==0.0.31-2005241907
+onnx
+ptflops

setup.py

@@ -0,0 +1,53 @@
+import os
+
+from setuptools import find_packages, setup
+
+version_file = "cover/version.py"
+
+
+def readme():
+    with open("README.md", encoding="utf-8") as f:
+        content = f.read()
+    return content
+
+
+def get_version():
+    with open(version_file, "r") as f:
+        exec(compile(f.read(), version_file, "exec"))
+    return locals()["__version__"]
+
+
+def get_requirements(filename="requirements.txt"):
+    here = os.path.dirname(os.path.realpath(__file__))
+    with open(os.path.join(here, filename), "r") as f:
+        requires = [line.replace("\n", "") for line in f.readlines()]
+    return requires
+
+
+setup(
+    name="cover",
+    version=get_version(),
+    description="Disentangled Video Quality Evaluator",
+    long_description=readme(),
+    long_description_content_type="text/markdown",
+    author="Teo (Timothy) Wu Hao Ning",
+    author_email="realtimothyhwu@gmail.com",
+    keywords="computer vision, video quality assessment",
+    url="https://github.com/twowu/cover",
+    include_package_data=True,
+    packages=find_packages(exclude=("examplar_data_labels", "figs")),
+    classifiers=[
+        "Development Status :: 4 - Beta",
+        "License :: OSI Approved :: Apache Software License",
+        "Operating System :: OS Independent",
+        "Programming Language :: Python :: 3",
+        "Programming Language :: Python :: 3.7",
+        "Programming Language :: Python :: 3.8",
+    ],
+    license="MIT License",
+    setup_requires=["numpy"],
+    install_requires=get_requirements(),
+    ext_modules=[],
+    cmdclass={},
+    zip_safe=False,
+)

train_one_dataset.py

@@ -0,0 +1,616 @@
+import torch
+
+import cv2
+import random
+import os.path as osp
+import argparse
+from scipy.stats import spearmanr, pearsonr
+from scipy.stats.stats import kendalltau as kendallr
+import numpy as np
+from time import time
+from tqdm import tqdm
+import pickle
+import math
+import wandb
+import yaml
+from collections import OrderedDict
+
+from functools import reduce
+from thop import profile
+import copy
+
+
+import cover.models as models
+import cover.datasets as datasets
+
+
+def train_test_split(dataset_path, ann_file, ratio=0.8, seed=42):
+    random.seed(seed)
+    print(seed)
+    video_infos = []
+    with open(ann_file, "r") as fin:
+        for line in fin.readlines():
+            line_split = line.strip().split(",")
+            filename, _, _, label = line_split
+            label = float(label)
+            filename = osp.join(dataset_path, filename)
+            video_infos.append(dict(filename=filename, label=label))
+    random.shuffle(video_infos)
+    return (
+        video_infos[: int(ratio * len(video_infos))],
+        video_infos[int(ratio * len(video_infos)) :],
+    )
+
+
+def rank_loss(y_pred, y):
+    ranking_loss = torch.nn.functional.relu(
+        (y_pred - y_pred.t()) * torch.sign((y.t() - y))
+    )
+    scale = 1 + torch.max(ranking_loss)
+    return (
+        torch.sum(ranking_loss) / y_pred.shape[0] / (y_pred.shape[0] - 1) / scale
+    ).float()
+
+
+def gaussian(y, eps=1e-8):
+    return (y - y.mean()) / (y.std() + 1e-8)
+
+
+def plcc_loss(y_pred, y):
+    sigma_hat, m_hat = torch.std_mean(y_pred, unbiased=False)
+    y_pred = (y_pred - m_hat) / (sigma_hat + 1e-8)
+    sigma, m = torch.std_mean(y, unbiased=False)
+    y = (y - m) / (sigma + 1e-8)
+    loss0 = torch.nn.functional.mse_loss(y_pred, y) / 4
+    rho = torch.mean(y_pred * y)
+    loss1 = torch.nn.functional.mse_loss(rho * y_pred, y) / 4
+    return ((loss0 + loss1) / 2).float()
+
+
+def rescaled_l2_loss(y_pred, y):
+    y_pred_rs = (y_pred - y_pred.mean()) / y_pred.std()
+    y_rs = (y - y.mean()) / (y.std() + eps)
+    return torch.nn.functional.mse_loss(y_pred_rs, y_rs)
+
+
+def rplcc_loss(y_pred, y, eps=1e-8):
+    ## Literally (1 - PLCC) / 2
+    y_pred, y = gaussian(y_pred), gaussian(y)
+    cov = torch.sum(y_pred * y) / y_pred.shape[0]
+    # std = (torch.std(y_pred) + eps) * (torch.std(y) + eps)
+    return (1 - cov) / 2
+
+
+def self_similarity_loss(f, f_hat, f_hat_detach=False):
+    if f_hat_detach:
+        f_hat = f_hat.detach()
+    return 1 - torch.nn.functional.cosine_similarity(f, f_hat, dim=1).mean()
+
+
+def contrastive_similarity_loss(f, f_hat, f_hat_detach=False, eps=1e-8):
+    if f_hat_detach:
+        f_hat = f_hat.detach()
+    intra_similarity = torch.nn.functional.cosine_similarity(f, f_hat, dim=1).mean()
+    cross_similarity = torch.nn.functional.cosine_similarity(f, f_hat, dim=0).mean()
+    return (1 - intra_similarity) / (1 - cross_similarity + eps)
+
+
+def rescale(pr, gt=None):
+    if gt is None:
+        pr = (pr - np.mean(pr)) / np.std(pr)
+    else:
+        pr = ((pr - np.mean(pr)) / np.std(pr)) * np.std(gt) + np.mean(gt)
+    return pr
+
+sample_types = ["semantic", "technical", "aesthetic"]
+
+
+def finetune_epoch(
+    ft_loader,
+    model,
+    model_ema,
+    optimizer,
+    scheduler,
+    device,
+    epoch=-1,
+    need_upsampled=False,
+    need_feat=False,
+    need_fused=False,
+    need_separate_sup=True,
+):
+    model.train()
+    for i, data in enumerate(tqdm(ft_loader, desc=f"Training in epoch {epoch}")):
+        optimizer.zero_grad()
+        video = {}
+        for key in sample_types:
+            if key in data:
+                video[key] = data[key].to(device)
+
+        y = data["gt_label"].float().detach().to(device).unsqueeze(-1)
+
+        scores = model(video, inference=False, reduce_scores=False)
+        if len(scores) > 1:
+            y_pred = reduce(lambda x, y: x + y, scores)
+        else:
+            y_pred = scores[0]
+        y_pred = y_pred.mean((-3, -2, -1))
+
+        frame_inds = data["frame_inds"]
+
+
+        loss = 0  # p_loss + 0.3 * r_loss
+
+        if need_separate_sup:
+            p_loss_a = plcc_loss(scores[0].mean((-3, -2, -1)), y)
+            p_loss_b = plcc_loss(scores[1].mean((-3, -2, -1)), y)
+            p_loss_c = plcc_loss(scores[2].mean((-3, -2, -1)), y)
+            r_loss_a = rank_loss(scores[0].mean((-3, -2, -1)), y)
+            r_loss_b = rank_loss(scores[1].mean((-3, -2, -1)), y)
+            r_loss_c = rank_loss(scores[2].mean((-3, -2, -1)), y)
+            loss += (
+                p_loss_a + p_loss_b + p_loss_c + 0.3 * r_loss_a + 0.3 * r_loss_b + 0.3 * r_loss_c
+            )  # + 0.2 * o_loss
+            wandb.log(
+                {
+                    "train/plcc_loss_a": p_loss_a.item(),
+                    "train/plcc_loss_b": p_loss_b.item(),
+                    "train/plcc_loss_c": p_loss_c.item(),
+                }
+            )
+
+        wandb.log(
+            {"train/total_loss": loss.item(),}
+        )
+
+        loss.backward()
+        optimizer.step()
+        scheduler.step()
+
+        # ft_loader.dataset.refresh_hypers()
+
+        if model_ema is not None:
+            model_params = dict(model.named_parameters())
+            model_ema_params = dict(model_ema.named_parameters())
+            for k in model_params.keys():
+                model_ema_params[k].data.mul_(0.999).add_(
+                    model_params[k].data, alpha=1 - 0.999
+                )
+    model.eval()
+
+
+def profile_inference(inf_set, model, device):
+    video = {}
+    data = inf_set[0]
+    for key in sample_types:
+        if key in data:
+            video[key] = data[key].to(device).unsqueeze(0)
+    with torch.no_grad():
+
+        flops, params = profile(model, (video,))
+    print(
+        f"The FLOps of the Variant is {flops/1e9:.1f}G, with Params {params/1e6:.2f}M."
+    )
+
+
+def inference_set(
+    inf_loader,
+    model,
+    device,
+    best_,
+    save_model=False,
+    suffix="s",
+    save_name="divide",
+    save_type="head",
+):
+
+    results = []
+
+    best_s, best_p, best_k, best_r = best_
+
+    for i, data in enumerate(tqdm(inf_loader, desc="Validating")):
+        result = dict()
+        video, video_up = {}, {}
+        for key in sample_types:
+            if key in data:
+                video[key] = data[key].to(device)
+                ## Reshape into clips
+                b, c, t, h, w = video[key].shape
+                video[key] = (
+                    video[key]
+                    .reshape(
+                        b, c, data["num_clips"][key], t // data["num_clips"][key], h, w
+                    )
+                    .permute(0, 2, 1, 3, 4, 5)
+                    .reshape(
+                        b * data["num_clips"][key], c, t // data["num_clips"][key], h, w
+                    )
+                )
+            if key + "_up" in data:
+                video_up[key] = data[key + "_up"].to(device)
+                ## Reshape into clips
+                b, c, t, h, w = video_up[key].shape
+                video_up[key] = (
+                    video_up[key]
+                    .reshape(b, c, data["num_clips"], t // data["num_clips"], h, w)
+                    .permute(0, 2, 1, 3, 4, 5)
+                    .reshape(b * data["num_clips"], c, t // data["num_clips"], h, w)
+                )
+            # .unsqueeze(0)
+        with torch.no_grad():
+            result["pr_labels"] = model(video, reduce_scores=True).cpu().numpy()
+            if len(list(video_up.keys())) > 0:
+                result["pr_labels_up"] = model(video_up).cpu().numpy()
+
+        result["gt_label"] = data["gt_label"].item()
+        del video, video_up
+        results.append(result)
+
+    ## generate the demo video for video quality localization
+    gt_labels = [r["gt_label"] for r in results]
+    pr_labels = [np.mean(r["pr_labels"][:]) for r in results]
+    pr_labels = rescale(pr_labels, gt_labels)
+
+    s = spearmanr(gt_labels, pr_labels)[0]
+    p = pearsonr(gt_labels, pr_labels)[0]
+    k = kendallr(gt_labels, pr_labels)[0]
+    r = np.sqrt(((gt_labels - pr_labels) ** 2).mean())
+
+    wandb.log(
+        {
+            f"val_{suffix}/SRCC-{suffix}": s,
+            f"val_{suffix}/PLCC-{suffix}": p,
+            f"val_{suffix}/KRCC-{suffix}": k,
+            f"val_{suffix}/RMSE-{suffix}": r,
+        }
+    )
+
+    del results, result  # , video, video_up
+    torch.cuda.empty_cache()
+
+    if s + p > best_s + best_p and save_model:
+        state_dict = model.state_dict()
+
+        if save_type == "head":
+            head_state_dict = OrderedDict()
+            for key, v in state_dict.items():
+                if "backbone" in key:
+                    continue
+                else:
+                    head_state_dict[key] = v
+            print("Following keys are saved :", head_state_dict.keys())
+            torch.save(
+                {"state_dict": head_state_dict, "validation_results": best_,},
+                f"pretrained_weights/{save_name}_{suffix}_finetuned.pth",
+            )
+        else:
+            torch.save(
+                {"state_dict": state_dict, "validation_results": best_,},
+                f"pretrained_weights/{save_name}_{suffix}_finetuned.pth",
+            )
+
+    best_s, best_p, best_k, best_r = (
+        max(best_s, s),
+        max(best_p, p),
+        max(best_k, k),
+        min(best_r, r),
+    )
+
+    wandb.log(
+        {
+            f"val_{suffix}/best_SRCC-{suffix}": best_s,
+            f"val_{suffix}/best_PLCC-{suffix}": best_p,
+            f"val_{suffix}/best_KRCC-{suffix}": best_k,
+            f"val_{suffix}/best_RMSE-{suffix}": best_r,
+        }
+    )
+
+    print(
+        f"For {len(inf_loader)} videos, \nthe accuracy of the model: [{suffix}] is as follows:\n  SROCC: {s:.4f} best: {best_s:.4f} \n  PLCC:  {p:.4f} best: {best_p:.4f}  \n  KROCC: {k:.4f} best: {best_k:.4f} \n  RMSE:  {r:.4f} best: {best_r:.4f}."
+    )
+
+    return best_s, best_p, best_k, best_r
+
+    # torch.save(results, f'{args.save_dir}/results_{dataset.lower()}_s{32}*{32}_ens{args.famount}.pkl')
+
+
+def main():
+
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "-o", "--opt", type=str, default="cover.yml", help="the option file"
+    )
+
+    parser.add_argument(
+        "-t", "--target_set", type=str, default="val-kv1k", help="target_set"
+    )
+
+    parser.add_argument('-n', "--name", type=str, default="COVER_TMP", help='model name to save checkpoint')
+    parser.add_argument('-uh', "--usehead", type=int, default=0, help='wheather to load header weight from checkpoint')
+
+    args = parser.parse_args()
+    with open(args.opt, "r") as f:
+        opt = yaml.safe_load(f)
+    print(opt)
+
+    ## adaptively choose the device
+
+    device = "cuda" if torch.cuda.is_available() else "cpu"
+
+    ## defining model and loading checkpoint
+
+    bests_ = []
+
+    if opt.get("split_seed", -1) > 0:
+        num_splits = 10
+    else:
+        num_splits = 1
+
+    print(opt["split_seed"])
+
+    for split in range(10):
+        model = getattr(models, opt["model"]["type"])(**opt["model"]["args"]).to(device)
+        if opt.get("split_seed", -1) > 0:
+            opt["data"]["train"] = copy.deepcopy(opt["data"][args.target_set])
+            opt["data"]["eval"] = copy.deepcopy(opt["data"][args.target_set])
+
+            split_duo = train_test_split(
+                opt["data"][args.target_set]["args"]["data_prefix"],
+                opt["data"][args.target_set]["args"]["anno_file"],
+                seed=opt["split_seed"] * (split + 1),
+            )
+            (
+                opt["data"]["train"]["args"]["anno_file"],
+                opt["data"]["eval"]["args"]["anno_file"],
+            ) = split_duo
+            opt["data"]["train"]["args"]["sample_types"]["technical"]["num_clips"] = 1
+
+        train_datasets = {}
+        for key in opt["data"]:
+            if key.startswith("train"):
+                train_dataset = getattr(datasets, opt["data"][key]["type"])(
+                    opt["data"][key]["args"]
+                )
+                train_datasets[key] = train_dataset
+                print(len(train_dataset.video_infos))
+
+        train_loaders = {}
+        for key, train_dataset in train_datasets.items():
+            train_loaders[key] = torch.utils.data.DataLoader(
+                train_dataset,
+                batch_size=opt["batch_size"],
+                num_workers=opt["num_workers"],
+                shuffle=True,
+            )
+
+        val_datasets = {}
+        for key in opt["data"]:
+            if key.startswith("eval"):
+                val_dataset = getattr(datasets, opt["data"][key]["type"])(
+                    opt["data"][key]["args"]
+                )
+                print(len(val_dataset.video_infos))
+                val_datasets[key] = val_dataset
+
+        val_loaders = {}
+        for key, val_dataset in val_datasets.items():
+            val_loaders[key] = torch.utils.data.DataLoader(
+                val_dataset,
+                batch_size=1,
+                num_workers=opt["num_workers"],
+                pin_memory=True,
+            )
+
+        run = wandb.init(
+            project=opt["wandb"]["project_name"],
+            name=opt["name"] + f"_target_{args.target_set}_split_{split}"
+            if num_splits > 1
+            else opt["name"],
+            reinit=True,
+            settings=wandb.Settings(start_method="thread"),
+        )
+
+        state_dict = torch.load(opt["test_load_path"], map_location=device)
+
+        # Load fine_tuned header from checkpoint
+        if args.usehead:
+            state_dict_head = torch.load(opt["test_load_header_path"], map_location=device)
+            for key in state_dict_head['state_dict'].keys():
+                state_dict[key] = state_dict_head['state_dict'][key]
+
+        # Allowing empty head weight
+        model.load_state_dict(state_dict, strict=False)
+
+        if opt["ema"]:
+            from copy import deepcopy
+
+            model_ema = deepcopy(model)
+        else:
+            model_ema = None
+
+        # profile_inference(val_dataset, model, device)
+
+        # finetune the model
+
+        param_groups = []
+
+        for key, value in dict(model.named_children()).items():
+            if "backbone" in key:
+                param_groups += [
+                    {
+                        "params": value.parameters(),
+                        "lr": opt["optimizer"]["lr"]
+                        * opt["optimizer"]["backbone_lr_mult"],
+                    }
+                ]
+            else:
+                param_groups += [
+                    {"params": value.parameters(), "lr": opt["optimizer"]["lr"]}
+                ]
+
+        optimizer = torch.optim.AdamW(
+            lr=opt["optimizer"]["lr"],
+            params=param_groups,
+            weight_decay=opt["optimizer"]["wd"],
+        )
+        warmup_iter = 0
+        for train_loader in train_loaders.values():
+            warmup_iter += int(opt["warmup_epochs"] * len(train_loader))
+        max_iter = int((opt["num_epochs"] + opt["l_num_epochs"]) * len(train_loader))
+        lr_lambda = (
+            lambda cur_iter: cur_iter / warmup_iter
+            if cur_iter <= warmup_iter
+            else 0.5 * (1 + math.cos(math.pi * (cur_iter - warmup_iter) / max_iter))
+        )
+
+        scheduler = torch.optim.lr_scheduler.LambdaLR(optimizer, lr_lambda=lr_lambda,)
+
+        bests = {}
+        bests_n = {}
+        for key in val_loaders:
+            bests[key] = -1, -1, -1, 1000
+            bests_n[key] = -1, -1, -1, 1000
+
+        for key, value in dict(model.named_children()).items():
+            if "backbone" in key:
+                for param in value.parameters():
+                    param.requires_grad = False
+
+        for epoch in range(opt["l_num_epochs"]):
+            print(f"Linear Epoch {epoch}:")
+            for key, train_loader in train_loaders.items():
+                finetune_epoch(
+                    train_loader,
+                    model,
+                    model_ema,
+                    optimizer,
+                    scheduler,
+                    device,
+                    epoch,
+                    opt.get("need_upsampled", False),
+                    opt.get("need_feat", False),
+                    opt.get("need_fused", False),
+                )
+            for key in val_loaders:
+                bests[key] = inference_set(
+                    val_loaders[key],
+                    model_ema if model_ema is not None else model,
+                    device,
+                    bests[key],
+                    save_model=opt["save_model"],
+                    save_name=args.name + "_head_" + args.target_set + f"_{split}",
+                    suffix=key + "_s",
+                )
+                if model_ema is not None:
+                    bests_n[key] = inference_set(
+                        val_loaders[key],
+                        model,
+                        device,
+                        bests_n[key],
+                        save_model=opt["save_model"],
+                        save_name=args.name
+                        + "_head_"
+                        + args.target_set
+                        + f"_{split}",
+                        suffix=key + "_n",
+                    )
+                else:
+                    bests_n[key] = bests[key]
+
+        if opt["l_num_epochs"] >= 0:
+            for key in val_loaders:
+                print(
+                    f"""For the linear transfer process on {key} with {len(val_loaders[key])} videos,
+                    the best validation accuracy of the model-s is as follows:
+                    SROCC: {bests[key][0]:.4f}
+                    PLCC:  {bests[key][1]:.4f}
+                    KROCC: {bests[key][2]:.4f}
+                    RMSE:  {bests[key][3]:.4f}."""
+                )
+
+                print(
+                    f"""For the linear transfer process on {key} with {len(val_loaders[key])} videos,
+                    the best validation accuracy of the model-n is as follows:
+                    SROCC: {bests_n[key][0]:.4f}
+                    PLCC:  {bests_n[key][1]:.4f}
+                    KROCC: {bests_n[key][2]:.4f}
+                    RMSE:  {bests_n[key][3]:.4f}."""
+                )
+
+        for key, value in dict(model.named_children()).items():
+            if "backbone" in key:
+                for param in value.parameters():
+                    param.requires_grad = True
+
+        for epoch in range(opt["num_epochs"]):
+            print(f"End-to-end Epoch {epoch}:")
+            for key, train_loader in train_loaders.items():
+                finetune_epoch(
+                    train_loader,
+                    model,
+                    model_ema,
+                    optimizer,
+                    scheduler,
+                    device,
+                    epoch,
+                    opt.get("need_upsampled", False),
+                    opt.get("need_feat", False),
+                    opt.get("need_fused", False),
+                )
+            for key in val_loaders:
+                bests[key] = inference_set(
+                    val_loaders[key],
+                    model_ema if model_ema is not None else model,
+                    device,
+                    bests[key],
+                    save_model=opt["save_model"],
+                    save_name=args.name + "_head_" + args.target_set + f"_{split}",
+                    suffix=key + "_s",
+                    save_type="full",
+                )
+                if model_ema is not None:
+                    bests_n[key] = inference_set(
+                        val_loaders[key],
+                        model,
+                        device,
+                        bests_n[key],
+                        save_model=opt["save_model"],
+                        save_name=args.name
+                        + "_head_"
+                        + args.target_set
+                        + f"_{split}",
+                        suffix=key + "_n",
+                        save_type="full",
+                    )
+                else:
+                    bests_n[key] = bests[key]
+
+        if opt["num_epochs"] >= 0:
+            for key in val_loaders:
+                print(
+                    f"""For the end-to-end transfer process on {key} with {len(val_loaders[key])} videos,
+                    the best validation accuracy of the model-s is as follows:
+                    SROCC: {bests[key][0]:.4f}
+                    PLCC:  {bests[key][1]:.4f}
+                    KROCC: {bests[key][2]:.4f}
+                    RMSE:  {bests[key][3]:.4f}."""
+                )
+
+                print(
+                    f"""For the end-to-end transfer process on {key} with {len(val_loaders[key])} videos,
+                    the best validation accuracy of the model-n is as follows:
+                    SROCC: {bests_n[key][0]:.4f}
+                    PLCC:  {bests_n[key][1]:.4f}
+                    KROCC: {bests_n[key][2]:.4f}
+                    RMSE:  {bests_n[key][3]:.4f}."""
+                )
+
+        for key, value in dict(model.named_children()).items():
+            if "backbone" in key:
+                for param in value.parameters():
+                    param.requires_grad = True
+
+        run.finish()
+
+
+if __name__ == "__main__":
+    main()